JP7175489B2 - BEAMFORMING METHOD, MEASUREMENT IMAGING DEVICE, AND COMMUNICATION DEVICE - Google Patents

Description

Application of Article 30, Paragraph 2 of the Patent Law IEICE Technical Research Report, IEICE Technical Report, Ultrasonics Vol. 116, No. 61, US2016-12-US2016-18, p. 25-30, US2016-16 (published on May 16, 2016, published by: The Institute of Electronics, Information and Communication Engineers). IEICE Technical Research Report, IEICE Technical Report, Ultrasonics Vol. 116, No. 190, US2016-38-US2016-43, p. 1-6, US2016-38 (issued on August 18, 2016, published by: The Institute of Electronics, Information and Communication Engineers). IEICE Technical Research Report, IEICE Technical Report, Ultrasonics Vol. 117, No. 227, US2017-56-US2017-63, p. 11-16, US2017-58 (published on September 28, 2017, published by The Institute of Electronics, Information and Communication Engineers).

The present invention relates to a beamforming method for beamforming using arbitrary waves arriving from a measurement object. Furthermore, the present invention relates to metrology imaging and communication devices using such beamforming methods.

In particular, the present invention captures an image of a measurement target based on arbitrary waves such as electromagnetic waves, light, mechanical vibrations, sound waves, or heat waves arriving from the measurement target, or measures the temperature, displacement, etc. of the measurement target. It relates to a digital beam forming method in an imaging device for non-destructively measuring the physical quantity, composition, structure, etc. of Objects to be measured include organic substances, inorganic substances, solids, liquids, gases, rheological objects, living creatures, celestial bodies, the earth, the environment, etc., and the range of applications is extremely wide.

The present invention is applicable to non-destructive inspection, diagnosis, resource exploration, generation and manufacture of materials and structures, monitoring of various physical or chemical repairs and treatments, and application of clarified functions and physical properties. In relation to these, there are cases where accuracy is required under conditions such as non-invasiveness, minimally invasiveness, and non-invasiveness, which do not cause great disturbance to the measurement object. Ideally, the object to be measured should be observed in situ.

In addition, the action of the waves itself may be used to treat or repair the object to be measured, and the response from the object to be measured at that time may be beam-formed to observe the situation. In addition, beamforming is implemented in satellite communications, radar, sonar, etc. to realize an informationally safe environment under energy saving, and accurate communications are being carried out. Beamforming has come to be applied to ad-hoc communication equipment and mobile communication as well. When the object to be measured is dynamic, real-time performance is required, and beamforming is required to be completed in a short time.

Waves such as electromagnetic waves, light, mechanical vibrations, sound waves, and heat waves behave differently depending on their frequency, bandwidth, intensity, or mode. Many types of wave transducers have been developed so far, and imaging is performed using transmitted waves, reflected waves, refracted waves, or scattered waves (forward scattered waves, backward scattered waves, etc.) of these waves.

For example, it is well known that ultrasound, which has a high frequency among sound waves, is used in non-destructive inspection, medical treatment, and sonar. Radar also uses electromagnetic waves (radio waves such as radio waves, FM waves, microwaves, terahertz waves, infrared rays, visible light, ultraviolet rays, X-rays, etc.) of appropriate frequencies according to the object to be measured. The same is true for other waves, and each wave behaves differently depending on its frequency. , polarization can also be used).

These applications often involve mechanically scanning the object to be measured with a transducer, using the same transducer multiple times, or using multiple pre-arrayed transducers for beamforming. . It is well known that aperture synthesis processing is performed when the earth, land, oceans, and weather are observed by satellites and airplane radars. Beamforming is often aimed at providing appropriate directivity and obtaining high spatial resolution and high contrast in a region of interest or a point of interest, particularly when imaging a measurement target.

As a result, by acting on the waves generated from the measurement object, we can observe reflection, transmission, various scattering (Rayleigh scattering, Mie scattering, etc.), attenuation, or their frequency dispersion caused by spatial changes in impedance. It is possible to observe not only what the measurement target is, but also the structure and composition of the interior and surface of the measurement target. The measurement target may be observed at various spatial resolutions. Characterization can be at the level of structure or composition (eg, individual level, molecular level, atomic level, nuclear level, etc.).

For the purpose of high-precision high-resolution imaging, signal compression techniques have been used for a long time, and chirp wave techniques and encoding compression techniques are typically used. In some cases, such as ISAR (inverse synthetic aperture radar), super-resolution is performed by inverting the beam characteristics of the observed wave motion (not necessarily during aperture synthesis). ), which may actively reduce the resolution. Among them, singular value decomposition, regularization, Wiener filter, etc. are effective.

Coding techniques are also used to separate multiple signals, such as separating a received signal into received signals corresponding to transmitted signals at different transmission locations. Waves coming from different directions may be separated, and the source of the signal may be identified and separated. In such cases, there is a great deal of reliance on matched filters with high signal detectability. However, although the signal energy can be obtained, the spatial resolution becomes low and the measurement accuracy decreases when the movement, displacement, strain, etc. of the object to be measured that accompanies deformation are to be obtained. The use of frequencies, bands, or multidimensional spectra is also useful in separating waves and signals.

In imaging using the above waves, the distribution of amplitude data usually obtained through quadrature detection, envelope detection, or square law detection is displayed in one, two, or three dimensions as a gray image or color image. and often become morphological images. Functional observations are also possible, eg in Doppler measurements using these waves, where raw coherent signals are processed (ultrasound Doppler, radar Doppler, laser Doppler, etc.).

In addition, for example, in the field of medical ultrasound, there is also a useful technique such as power Doppler that can detect moving tissue without information on the direction of displacement. Moreover, when microwaves, terahertz waves, or far infrared rays are used, the temperature distribution of the measurement target can also be observed. These measured physical quantities may be superimposed and displayed on the morphological image. In the field of image metrology, it is also well known that motion observations are made using signals made incoherent through detection (cross-correlation processing, optical flow, etc.). In medical ultrasound and sonar, imaging using physically generated harmonics, chords, and difference tones is also performed. In particular, when the object to be measured is dynamic, the beamforming process is required to be real-time.

Satellite communications, radar, sonar, etc. also implement beamforming to realize an informationally safe environment under energy conservation, and accurate communications are being carried out. Beamforming has come to be applied to ad-hoc communication equipment and mobile communication as well. It is also effective for communication with a specific person, a specific signal source, or a specific position. In communication, information is sent from the transmitting side to the receiving side in waves, and the purpose may be achieved by itself. Of course, the example of communication is not limited to this. When the content of information is dynamic according to communication targets and observation targets, real-time performance is required, and beamforming in this case is required to be completed in a short period of time.

Under such circumstances, the inventor of the present application is developing an ultrasonic imaging technique for differential diagnosis of lesions such as cancer lesions and sclerosis in human tissues. The inventor of the present application has improved the resolution and efficiency of HIFU (High Intensity Focus Ultrasound) treatment, along with the improvement of echo imaging resolution and high-precision measurement imaging of tissue displacement. We are also imaging them based on the reception of echoes during intense ultrasound radiation. These imaging techniques are based on high-speed and appropriate beamforming, and we have previously reported on high-speed and appropriate detection methods, tissue displacement measurement methods, and shear wave propagation measurement methods.

More than 20 years have passed since medical ultrasound diagnostic imaging equipment was digitized. Earlier, mechanical scanning was performed using a single aperture transducer element, followed by electronic scanning using multiple transducer elements or arrays of these transducers, and signal processing devices were known as analog devices. to digital equipment. In fact, the classical aperture synthesis process itself has been digital beamforming since it was first used in radars mounted on satellites and airplanes, but the strength of the received signal is weak. For this reason, it has been rarely used in medical ultrasound.

On the other hand, in recent years, the inventor of the present application has invented a multidirectional aperture synthesis method, and performed beamforming in multiple directions from received echo data for aperture synthesis. As a result, it is possible to obtain multi-directional steering (deflection) image signals at the same frame rate as that of normal electronic scanning. It has a transverse carrier frequency that is orthogonal to this, and a higher spatial resolution than normal imaging is possible. Furthermore, real-time measurement of the displacement vector distribution has become possible by using the multidimensional autocorrelation method, which was also invented by the inventor of the present application. In addition, in the case of incoherent addition, it is also possible to reduce speckle (usually, transmission beams are generated in different directions to reduce speckle, but according to the multidirectional aperture synthesis method, , which can achieve high frame rates). In addition to the field of ultrasonic waves, digitization is progressing in sensing devices that use microwaves, terahertz waves, radiation such as X-rays, other electromagnetic waves, vibration waves including sound, and waves such as heat waves for non-destructive inspection. It is

For example, aperture synthesis in these sensing devices is active beamforming, and the waves to be processed are transmitted, reflected, refracted, or scattered waves (forward scattered waves or backscattered waves, etc.). On the other hand, in passive beamforming, for example, the temperature distribution measurement based on the above-mentioned microwave observation and the electrical activity source by the brain magnetic field of living things are observed. Transmitted waves, reflected waves, refracted waves, or scattered waves (forward scattered waves, backward scattered waves, etc.) based on waves emitted from a (self-emanating) signal source are targeted. There are many other examples that correspond to these, but as another example, recently, a living organism is used as a measurement object, called photoacoustic, laser irradiation is performed on the measurement object, and frequency-dependent A volume change is caused by a certain amount of heat absorption, an ultrasonic wave is generated using this as a sound source, and peripheral blood vessels are discriminated as a result of beam forming.

Compared to analog devices, digital devices require more processing time, but their computational processing capabilities have improved dramatically, they have become smaller, they have become cheaper, including data storage media. There are many advantages such as easy application and high degree of freedom. In fact, even though it is a digital device, high-speed analog processing after receiving a signal by a sensor is extremely important, especially in a sensing device. ), it is appropriately realized in combination with the digital processing after being processed.

In analog devices, transmit and receive beamforming is performed by analog processing. On the other hand, in a digital device, transmit beamforming is done by analog or digital processing, and receive beamforming is done by digital processing. Therefore, in the present application, a digital beamformer means one that necessarily performs reception beamforming by digital processing.

After receiving the waves from the object to be measured through a plurality of conversion elements, an array device composed of them, or mechanical scanning of one or more conversion elements, DAS (Delay and Summation), which is a so-called aperture synthesis process, is performed. addition) processing is performed. In transmission, transmission beamforming may be performed by driving a plurality of elements, and classical aperture synthesis based on single-element transmission may be performed. , DAS processing is performed.

That is, transmit beamforming is performed by analog processing or digital processing. On the other hand, in receive beamforming, waves are received by each element in the array or elements at different positions to generate received signals, and after level adjustment by analog amplification or attenuation, analog filtering, etc., analog received signals are generated. A/D conversion is performed to a digital received signal, and the digital received signal is stored in a memory. After that, devices and calculators with general-purpose calculation processing capabilities, PLDs (Programmable Logic Devices), FPGAs (Field-Programmable Gate Arrays), DSPs (Digital Signal Processors), A GPU (Graphical Processing Unit), or a microprocessor, or a dedicated computer, a dedicated digital circuit, or a dedicated device is used to perform digital processing on the stored received signal. be.

These digital processing devices may also include these analog devices, AD converters, memories, and the like. Devices and computers with computing power may be multi-core. This makes it possible to easily perform dynamic focusing during reception, which is almost impossible with analog devices. Parallel processing is also performed. Transmission lines (for example, laminated circuits) and broadband radio paths are important for increasing the speed of analog processing and digital processing.

Dynamic focusing improves the spatial resolution in the range direction of the generated image signal and in the depth direction (depth direction) of the object to be measured. On the other hand, dynamic focusing of the transmission is possible only when classical aperture synthesis with one-element transmission. In order to secure the energy of the transmission wave, instead of aperture synthesis by single-element transmission, transmission beamforming at a fixed focal position by driving multiple elements is often performed.

The inventors of the present application have invented high frame rate echo imaging that transmits a wave with a wide lateral wavefront, such as a plane wave, and interrogates a large area in a single transmission. Furthermore, the inventors of the present application coherently add (superimpose) a plurality of waves of this kind with different steering (deflection) angles to achieve lateral modulation and lateral broadband (high lateral resolution). In particular, when the above-mentioned multidimensional autocorrelation method is realized, shear wave propagation, blood flow in the carotid artery with fast blood flow, blood flow in the heart chamber with complicated flow, etc. is measured as a multi-dimensional displacement vector. This is the same when multidirectional aperture synthesis is performed and when transmission beamforming is performed. In addition, waves having different carrier frequencies are generated and superimposed to achieve axial broadband (high resolution in the axial direction).

Such processing is performed in the case of active beamforming, but the transmitter is not used in passive beamforming. Thus, a digital beamformer usually consists of a transmitter (in the case of active beamforming), a receiver, and a DAS processing device, and is realized by assembling these devices. Small packages are now available at low cost.

In this DAS processing, for phasing, interpolation approximation processing is mixed in the spatial domain to apply a delay to the received signal at high speed. There is a method that applies a delay with high accuracy based on the Nyquist theorem (past inventions of the inventor of the present application , Patent Document 6, Non-Patent Document 15, etc. ), and adds received signals in the spatial domain after phasing. (phased addition). In a digital device, a command signal generated by a control unit may be used as a trigger signal for digital sampling (AD conversion) of an analog received signal, e.g. .

Also, when driving multiple elements with a transmission delay in one beamforming, an analog or digital transmission delay that is installed in the transmission unit in advance and realizes the transmission focus position, steering direction, etc. that can be selected by the operator. You may use patterns. Further, in the receiving digital processing, the command signal for the first driven element, the last driven element, or the command signal for driving another element is used as a trigger signal, In some cases, sampling is started for the signal of , and a digital delay is applied. These command signals may be generated based on a command signal for starting beamforming for one frame.

If a digital delay is implemented in the transmission delay, unlike an analog delay, an error determined by the clock frequency for generating the digital control signal will inevitably occur, so the analog delay is better for the transmission delay. In addition, if a digital delay is applied in reception to realize reception dynamic focusing, an error occurs in the above-described interpolation approximation process. There was no choice but to apply a precise digital delay (phase rotation processing) and use low-speed beamforming.

In the former case involving interpolation approximation, phasing addition may simply add echo signals from positions containing the coordinate positions of the signal values to be generated, or may be bilinear (bi-linear) or multi-dimensional polynomial Accuracy may be improved by interpolation using . The former is much faster but less accurate than the latter, which uses complex exponentials. The latter is highly accurate but extremely slow. This phased summation is performed under the assumption that the propagation velocity of the wave is known or assumed, and is often assumed to be constant within the region of interest. On the other hand, the propagation velocity is also measured and the phase aberration is corrected. Aberrations can be determined (if the velocity of propagation is homogeneous, then interferometry).

A two-dimensional or three-dimensional distribution of the aperture elements, or a two-dimensional or three-dimensional array of the aperture elements, as compared to a distribution or array of the aperture elements in one-dimensional space. In some cases, it is necessary to perform more processing for beamforming, and many processors are installed to perform parallel processing. Beamforming at distant locations with little interference, transmission beamforming in multiple directions with different steering angles, and single transmission beamforming may be processed in parallel reception.

In terms of communication control, it is necessary to properly generate waves that reflect the type and amount of communication data, or the characteristics of the medium. It is desirable that communication takes place. Interfering waves can also be separated using analog devices or by analog or digital signal processing. Controlled waves of propagation direction, coding, frequency and/or bandwidth are important.

As another invention similar to the multidirectional aperture plane synthesis processing described above, it is possible to improve the frame rate by performing multidirectional reception beamforming for one transmission beamforming. Also, in beamforming, apodization may be important. For example, transmit and receive apodization may each be implemented to reduce sidelobes, but this trades off lateral resolution and should be implemented appropriately. On the other hand, a simple beamformer without apodization is often used, emphasizing spatial resolution. However, the inventors of the present application report that appropriate apodization is necessary to suppress side lobes and obtain lateral resolution during steering. Also, some of the past inventions of the inventor of the present application remove side lobes in the frequency domain.

In addition, contrast agents (for example, microbubbles in medical ultrasound imaging) may be used in order to apply the nonlinear characteristics of waves in an object. irradiating high-intensity waves and waves containing harmonics (broadband transmission), and applying nonlinear processing to received coherent signals and coherent signals after phasing and addition, realizing high spatial resolution and side Invented high-contrast imaging by suppressing lobes. The inventors of the present application also invented highly accurate tissue displacement (vector) measurement based on the same nonlinear processing.

It is also possible to generate imaging signals using virtual sources. As for the virtual source, there have been reports in the past that a virtual source is installed in front of a physical aperture and that a virtual source is installed at the transmission focal position. Furthermore, the inventors of the present application have found that the detector as well as the virtual source can be placed at arbitrary positions without the use of a focal point, or that the physical sources and detectors of the waves can be placed in any suitable scatterer or diffraction grating. It is reported that it can be installed in the position. It is possible to increase spatial resolution and widen the field of vision (FOV).

Quadrature detection, envelope detection, and square-law detection are used as imaging forms, but the inventors of the present application actively display waviness itself as a color image or a gray image, and phase information is used. Emphasis is placed on displays that include Thus, multidimensional devices using various waves are being developed for various purposes.

There are several methods of digital beamforming using fast Fourier transforms that have been disclosed so far, one of which is analog processing ( Non-Patent Document 1) is digitized, and exactly classical aperture plane synthesis is performed at high speed and with high accuracy through fast Fourier transform (see Non-Patent Document 2). In this processing, interpolation approximation is not required. No digital processing is disclosed.

In addition, digital beamforming disclosed elsewhere requires conventional interpolation approximation processing using sampling data of nearby positions (paragraph 0197), bilinear, multidimensional polynomial, etc., and accuracy is low. . For example, a method of performing wave number matching through fast Fourier transform, including steering in plane wave transmission, has been disclosed (see Non-Patent Documents 3-5), but in that case the aperture shape of the array is not flat. When performing calculations and displaying images (when the array aperture is an arc (see Non-Patent Document 6)), conventional interpolation approximation processing is required and accuracy is low . Techniques using the fast Fourier transform during transmission of plane waves are also disclosed in Patent Documents 1 to 4, but all of them perform wave number matching through conventional interpolation approximation. However, by wavenumber matching by interpolation approximation, the wavenumber vector coordinate system of the angular spectrum is equally spaced to obtain a multidimensional spectrum, and by performing a fast inverse Fourier transform, beamforming can be completed at high speed. There are advantages .

A recent non-patent document 5 discloses wave number matching to Fourier transform for non-equidistant sampling signals, which is also based on conventional interpolation approximation. As mentioned above, digital beamforming already has a long history, but when real-time performance (computation speed) until displaying an image is most important, conventional interpolation approximation is often used, and the best accuracy is achieved. did not provide it. In addition, regarding the most popular fixed focus processing known to perform DAS processing and steering at that time, even a processing method through digital fast Fourier transform is not disclosed.

There is also a report on a migration method (see, for example, Non-Patent Document 7), which also requires conventional interpolation approximation in wave number matching .

U.S. Pat. No. 5,720,708 U.S. Pat. No. 6,685,641 U.S. Pat. No. 7,957,609 U.S. Patent Application Publication No. 2009/0036772 U.S. Pat. No. 8,211,019 U.S. Pat. No. 7,775,980 U.S. Patent Application Publication No. 2011/0172538 Japanese Patent No. 5441292 U.S. Pat. No. 9,326,748 U.S. Pat. No. 7,690,838 U.S. Pat. No. 9,084,559

J. W. Goodman, "Introduction to Fourier Optics" 2nd ed., McGraw-Hill Co, Inc., 1996 L. J. Busse, IEEE Trans. UFFC, vol. 39, no. 2, pp. 174-179, 1992 J. Cheng, J.-y. Lu, IEEE Trans UFFC, vol. 53, no. 5, pp. 880-899, 2006 H. Peng, J.-y. Lu, X. Han, Ultrasonics, 44, e97-e99, 2006 P. Kruizinga et al, IEEE Trans. UFFC, vol. 59, no. 12, pp. 2684-2691, 2012 M. A. Haun, D. L. Jones, W. D. O'Brien, Jr., IEEE Trans. UFFC, vol. 49, pp. 861-870, 2002 C. Sumi, IEEE Trans. UFFC, vol. 55, pp. 2607-2625, 2008 C. Sumi, S. Uga, "Effective ultrasonic virtual sources which can be positioned independently of physical aperture focus positions," Rep. Med. Imag., vol. 3, pp. 45-59, 2010 M. Soumekh, "Fourier array imaging," PTR Prentice-Hall, Englewood Cliffs, New Jersey 07632, 1994 S. Haykin, A. Steinhardt ed. "Adaptive radar detection and estimation," John Wiley & Sons, inc. New York. 1992 K.W. Hollman, K.W. Rigby, M. O'Donnell, "Coherence Factor of Speckle from a Multi-Row Probe," Proc. of IEEE Ultrasonics Symp, pp. 1257-1260, 1999 D. Garcia, L. L. Tarnec, S. Muth, E. Montagnon, J. Poree, G. Cloutier, IEEE Trans. UFFC , vol. 60, no. 9, pp. 1853-1867, 2013 C. Sumi, IEEE Trans. UFFC, vol.55, pp.24-43, 2008 C. Sumi, Rep. Med. Imag., vol. 5, pp. 57-101, 2012 C. Sumi, IEEE Trans. UFFC, vol.46, pp.158-166, 1999 S. Srinivasan, F. Kallel, and J. Ophir, Ultrasound Med. Biol., vol. 28, pp. 359-368, 2002 C. Sumi, IEEE Trans. UFFC, vol.55, pp.297-307, 2008 C. Sumi and K. Sato, IEEE Trans. UFFC, vol. 55, pp. 787-799, 2008 C. Sumi, Y. Takanashi, K. Ichimaru, Rep. Med. Imag., vol. 5, pp. 23-50, 2012 C. Kasai, K. Namekawa, A. Koyano, R. Omoto, "Real-time two-dimensional blood flow imaging using an autocorrelation technique," IEEE Trans. On Sonics and Ultrasonics, vol. 32, pp. 458-464, 1985 J. A. Jensen, "Field: A program for simulating ultrasound systems," Med, Biol, Eng, Comp, 10th Nordic-Baltic Conference on Biomedical Imaging, Vol. 4, Supplemental, Part 1: 351-353, 1996 B. Schrope, V. L. Newhouse, V. Uhlendorf, "Simulated capillary blood flow measurement using a nonlinear ultrasonic contrast agent," Ultrason. Imag., vol. 14, pp. 134-158, 1992 P. N. Burns, S. R. Wilson, D. H. Simpson, "Pulse Inversion Imaging of Liver Blood Flow: Improved Method for Characterizing Focal Masses with microbubble Contrast," Investigative Radiology, vol. 35, no. 1, p. 58, 2000 M. A. Averkiou, D. N. Roundhill, J. E. Powers, "A new imaging technique based on the nonlinear properties of tissues," 1997 IEEE Ultrasonics symp, pp. 1561-1566, 1997 B. Haider, R. Y. Chiao, "Higher order nonlinear ultrasonic imaging," 1999 IEEE Ultrasonics symp., pp. 1527-1531, 1999 A. Needles, M. Arditi, N. G. Rognin, J. Mehi, T. Coulthhard, C. Bilan-Tracey, E. Gaud, P. Frinking, D. Hirson, F. S. Foster, "Nonlinear contrast imaging with an array-based micro -ultrasound system," Ultrasound Med. Biol., vol. 36, no. 12, pp. 2097-2106, 2010 J. R. Doherty, G. E. Trahey, K. R. Nightingale, M. L. Palmeri, "Acoustic Radiation Force Elasticity Imaging in Diagnostic Ultrasound," IEEE Trans. on UFFC, vol. 60, no. 4, pp. 685-701, April 2013 K. Hynynen, "Demonstration of enhanced temperature elevation due to nonlinear propagation of focused ultrasound in dog's thigh in vivo," Ultrasound Med. Biol. vol. 13, no. 2, pp. 85-91, 1987 Y. Huang, N. I. Vykhodtseva, K. Hynynen, "Creating brain lesions with low-intensity focused ultrasound with mcrobubbles: A rat study at half a megahertz," Ultrasound Med. Biol., vol. 39, no. 8, pp. 1420 -1428, 2013 C. Sumi, "Utilization of an ultrasound beam steering angle for measurements of tissue displacement vector and lateral displacement," Reports in Medical Imaging, vol. 3, pp. 61-81, 2010 A. K. Katsaggelos, K. T. Lay, "Maximum likelihood blur identification and image restoration using the EM algorithm," IEEE Trans. Signal Processing, vol. 39, no. 3, pp. 729-733, 1991 R. Molina, A. K. Katsaggelos, J. Mateos, "Baysian and regularization methods for hyperparameter estimation in image restoration," IEEE Trans. Image Processing, vol. 8, no. 2, pp. 231-246, 1999 M. Nikolovia, "Markovian reconstruction using a GNC approach," IEEE Trans. Image Processing, vol. 8, no. 9, pp. 1204-1220, 1999 R. Molina, Javier Mateos, A. K. Katsaggelos, M. Vega, "Baysian multichannel image restoration using comound Gauss-Markov random fields," IEEE Trans. Image Processing, vol. 12, no. 12, pp. 1642-1654, 2003 Kokubo, Yagi, and Nakayama, "Improving the Resolution of Ultrasound Images Using a Two-Dimensional Complex Inverse Filter," Acoustical Society of Japan, vol. 47, no. 7, pp. 443-450, 1991 Morohoshi, Nakayama, and Yagi, "Characteristics evaluation of ultrasound echo diagnostic equipment based on two-dimensional AR model and improvement of image resolution," IEICE Transactions on Information and Systems, vol. 76(DII-2), pp. 233- 240, 1993 C. L. Chan, A. K. Katsaggelos, "Iterative maximum likelihood displacement field estimation in quantum-limited image sequence," IEEE Trans. Image Processing, vol. 4, no. 6, pp. 743-751, 1995 J. C. Brailean, A. K. Katsaggelos, "Simultaneous recursive displacement estimation and restoration of noisy-blurred image sequences," IEEE Trans. Image Processing, vol. 4, no. 9, pp. 1236-1251, 1995 Y.-L. You, M. Kaveh, "Blind Image restoration by anisotropic regularization," IEEE Trans. Image Processing, vol. 8, no. 3, pp. 396-407, 1999 T. Chan, C.-K. Wong, "Total variation blind deconvolution," IEEE Trans. Image Processing, vol. 7, no. 3, pp. 370-375, 1998 F. Sroubek, J. Flusser, "Multichannel blind iterative image restoration," IEEE Trans. Image Processing, vol. 12, no. 9, pp. 1094-1106, 2003 Miles N. Wernick et al, "Fast spatio-temporal image reconstruction for dynamic PET," IEEE Trans on Medical Imaging, vol. 18, pp. 185-195, 1999 A. K. Katsaggelos, J. Biemond, R. W. Schafer, R. M. Mersereau, "A regularized iterative image restoration algorithm," IEEE Trans Signal Processing, vol. 39, pp. 914-929, 1991

As described above, if a digital delay is applied in reception to achieve reception dynamic focusing, an error occurs in the interpolation approximation process described above. Alternatively, there was no choice but to apply the above-mentioned high-precision digital delay (phase rotation processing) and use low-speed beamforming.

So far, we have studied electromagnetic waves, vibrational waves (mechanical waves) including sound waves (compression waves), shear waves, surface waves, etc., and waves such as heat waves. Disclosed is a method of digital beamforming for waves such as scattered or backscattered waves, refracted waves, surface waves, shock waves, or those originating from a self-emanating wave source: As mentioned above, it is limited to monostatic aperture synthesis without steering, plane wave transmission with steering, and migration method. Also, other than its monostatic aperture synthesis, digital beamforming could not be implemented without the need for interpolation approximations. Therefore, its accuracy was low.

In contrast, when using a transmitting or receiving transducer array device with an arbitrary aperture shape (which may be used for both transmitting and receiving, or the transmitting and receiving may be on different waves), or when using passive When only the receiving transducer is used in such beamforming, conventional interpolation approximation is required regardless of the presence or absence of focusing or steering of transmission or reception, even if the coordinate system of the transmitted and received beams and the coordinate system for image display are different. It is desired to perform arbitrary beamforming at high speed and with high precision without using

In the case of active beamforming, transmit and receive transducer array devices (transducers may be used for both transmit and receive) with arbitrary aperture shapes are used. Also, for passive beamforming, only receive transducer array devices with arbitrary aperture shapes are used. In these cases, digital processing is desired to realize arbitrary beamforming at high speed and with high accuracy. It is desired to perform virtually arbitrary focusing and steering (deflection) with a transducer array device having an arbitrary aperture shape.

After beamforming with phasing and addition, linear or nonlinear signal processing is applied to a plurality of beams in which at least one of wave parameters such as frequency, bandwidth, pulse shape, beam shape, etc. in each direction is different, A beam with a new value for at least one of these wave parameters may be generated (frequency modulated, broadband, multifocus, etc.) , but in such beamforming , focusing and steering (deflection), and apodization have been performed using transducer array devices with arbitrary aperture shapes, usually based on DAS processes involving less accurate conventional interpolation approximations . As will be described later , linear or nonlinear phenomena in the medium may generate and exploit beams with new and different values for at least one of their wave parameters.

Since the wave propagation velocity is determined by the physical properties of the medium under physical conditions, when the aperture element configures a two-dimensional or three-dimensional distribution or a multi-dimensional array to perform imaging in a multi-dimensional space, the one-dimensional case Compared to , a lot of beamforming is performed, and the number of processing data used for one beamforming also increases, so it takes a huge amount of time. It is desirable to use a device that processes these beamforming processes in real time or a device that can display results in a short period of time.

Also, until now, regarding digital beamforming through Fourier transform, it has been disclosed that it is performed through conventional interpolation approximation in a Cartesian coordinate system when using a one-dimensional or two-dimensional linear array type transducer. It is desirable to perform digital beamforming without performing conventional interpolation approximation even when the coordinate system at the time of transmission or reception is different from that of the result (image) display.

The methods disclosed for cases where the array aperture shape is not flat (eg, where the array aperture is an arc) also provided conventional interpolation approximations . As a typical example, when using a convex transducer, or when performing electronic or mechanical sector scanning or IVUS (intravascular ultrasound) scanning, transmission and reception are performed in an arbitrary coordinate system such as a polar coordinate system. It would be desirable to process the digital signal obtained from the waves and obtain the beamformed signal directly in an arbitrary image display system (arbitrary coordinate system), such as the Cartesian coordinate system, without conventional interpolation approximation. be

In recent years, memory and AD converters have become very inexpensive, but by sampling waves based on the Nyquist theorem without oversampling, DAS processing that takes time when interpolation approximation is not performed It is desirable to be able to perform the corresponding high-precision beamforming at high speed. Proper execution of apodization can also be important.

As a result of solving these problems, the spatial resolution and contrast of image signals obtained in real time or in a short time (including the effect of suppressing side lobes) are high, and the movement (displacement) of the object from the obtained signals is improved. ), deformation, temperature, etc., it is desired to realize highly accurate measurement. For example, in the field of medical ultrasound, in recent years, after applying the Doppler method to echo signals to measure tissue displacement and velocity, this is subjected to spatio-temporal differentiation to obtain acceleration, strain, etc., and image it. became. Since spatio-temporal differentiation is a process that amplifies high-frequency measurement errors and degrades the SN ratio (Signal-to-Noise Ratio), it is necessary to achieve highly accurate displacement measurement using phase. , The high-precision beamforming that realizes this is based on conventional DAS processing (processing that does not include approximate calculation for adding phasing signals obtained by phase rotation processing in the frequency domain, Patent Document 6, Non-Patent Document 15, etc.) Based on dynamic focusing only . Three-dimensional imaging apparatuses using two-dimensional arrays and three-dimensional arrays are also becoming popular, so arbitrary beamforming including dynamic focusing can be performed at high speed and without conventional interpolation approximation. It is desirable to achieve this with high precision.

The inventor of the present application has recently developed a highly accurate measurement method for relatively fast-moving tissue displacement and shear wave propagation based on high-speed beamforming (high-speed transmission and reception within the region of interest) using polarized plane wave transmission. However, even when such focusing is not performed, it is desirable to perform high-speed beamforming with high accuracy without interpolation approximation. By performing high-speed beamforming while changing the steering angle and performing coherent summation, it is possible to obtain almost the same high image quality (spatial resolution and contrast) at high speed as compared to scanning using a normal focusing beam. . Fast beamforming is also effective for multidimensional imaging using multidimensional arrays.

In addition, steering in classical aperture plane synthesis (monostatic type) based on scanning by driving one element at a time and multi-static aperture plane synthesis, which has not been disclosed so far, can be performed at high speed and without interpolation approximation. It is desirable to carry out with a high degree of accuracy. Also, even when so-called migration processing is used, it is desirable to process arbitrary beamforming at high speed and with high accuracy without performing interpolation approximation in an arbitrary coordinate system. Other specific examples of beamforming that are desired to be implemented are as described elsewhere in this specification, and it is desired that such beamforming be similarly performed with high speed and high accuracy. be

Accordingly, one of the objects of the present invention is to use a digital beamformer having a digital operation function and perform arbitrary beamforming at high speed and with high accuracy without performing conventional approximate calculation. This enables various applications of waves described below, including super-resolution using non-linear processing. There are various other applications such as imaging, displacement measurement, and temperature measurement. High-precision and high-speed Hilbert transform processing using a transform (a multidimensional Fourier transform is performed on the received multidimensional signal, and through zero padding of the spectrum in the frequency domain, an octant spectrum in the case of three dimensions, and an octant spectrum in the case of two dimensions). It is to realize a process that is faster or computationally simpler than the method of generating a quadrant spectrum and performing a multi-dimensional inverse Fourier transform in the case of .

The present invention has been made to solve at least part of the above problems. A measurement imaging apparatus according to one aspect of the present invention is configured such that, in a rectangular coordinate system defined by a receiving aperture element array, when arbitrary waves are transmitted from a wave source located in an arbitrary direction toward a measurement target, (1) a plurality of reception aperture elements that receive waves arriving from the object to be measured and generate respective reception signals; (2) the reception signals ; (a) applying beamforming processing to one of a quadrature detection signal of a received signal, an IQ signal (in-phase signal and quadrature signal) of the received signal, and an analysis signal of the received signal; phase rotation based on multiplication of a complex exponential function whose exponent part is the product of a selected single angular frequency at a sampling instant near the instant of and the time difference between said desired instant and said sampling instant or (b) a selected single angular wavenumber at a spatial sampling location near a desired spatial location to which the beamforming process is applied, and between said desired spatial location and said spatial sampling location. obtaining the signal at the desired point in time or spatial position by interpolation approximation by applying at least a phase rotation based on multiplication of a complex exponential function having in the exponent part the product of the distance and the distance of the plurality of receive aperture elements, phasing and addition processing is performed using a plurality of signals, and a transmission component or a signal processing unit that generates the image signal in a desired coordinate system having coordinate axes in at least two directions by performing beamforming processing for transmission or reception including steering for reception. Here, the wavenumber is the spatial frequency of the wave motion, and the wavenumber of a sine wave is defined as the reciprocal of the wavelength or its 2π times, the latter being particularly called the angular wavenumber. Also, the steering angle is represented by a declination angle in a two-dimensional orthogonal coordinate system, and represented by an elevation angle and an azimuth angle in a three-dimensional orthogonal coordinate system.

さらに、その積を、軸方向ｙに関して分解能を持たせるべく横方向ｘに関してフーリエ変換し、得られた演算結果に、横方向ｘに行われた波数マッチングの効果を除いて実施できる複素指数関数（１０２）を掛けると同時に、複素指数関数（１０３）を掛けることにより軸方向に関する波数マッチングを行い、ここで、横方向の波数がｋ_ｘと表され、

補間近似処理を行うことなく波数マッチングを行うことによって、直接的にデカルト座標系においてイメージ信号を生成することを含む。
また、本発明の他の観点に係るビームフォーミング方法は、フーリエビームフォーミングやＤＡＳ処理等他、様々なビームフォーミングにおいて、アレイ型開口素子群（各々の素子が独立に駆動され、また、独立に受信信号を得ることのできる、独立した送信又は受信のチャンネルを持つ）において、隣接する又は離れた位置の素子に同一の送信又は受信の遅延や送信又は受信のアポダイゼーションを施して１つの開口として使用することにより、１素子を用いた送信又は受信よりも強さの大きい波動を送信又は受信のビームフォーミングに使用することを含む。 Further, a beamforming method according to another aspect of the present invention is a Cartesian orthogonal coordinate system using coordinates of an axial direction y determined by the orientation of apertures of a flat receive aperture element array and a lateral direction x perpendicular to the axial direction y. Arbitrary waves are transmitted from the wave source located at to the object to be measured, and the waves coming from the object to be measured are transmitted in a direction where the deflection angle (deviation angle) θ formed with the axial direction is zero degrees or non-zero degrees, or processed as received, and furthermore, when the same wave arriving from the measurement object is received and dynamically focused in a direction in which the deflection angle (deviation angle) φ formed with the axial direction is zero or non-zero degrees, the measurement object Step (a) of receiving waves coming from an object by at least one receive aperture element to generate a received signal; and performing at least Fourier transform and wave number matching on the received signal generated in step (a). and step (b) of performing beamforming processing, wherein step (b) is performed on the received signal without performing wavenumber matching including interpolation approximation processing in the wavenumber domain or the frequency domain on the received signal. The complex _{exponential function} ( ₁₀₁ ) to perform wave number matching in the horizontal direction x,

Furthermore, the product is subjected to Fourier transform in the horizontal direction x to provide resolution in the axial direction y, and the obtained operation result is a complex exponential function ( 102) and simultaneously multiply by the complex exponential function (103) to perform wavenumber matching in the axial direction, where the transverse _wavenumber is denoted as kx,

It involves generating an image signal in the Cartesian coordinate system directly by wavenumber matching without interpolation approximation.
In addition, a beamforming method according to another aspect of the present invention is an array aperture element group (each element is independently driven and independently received) in various beamforming such as Fourier beamforming and DAS processing. have independent transmit or receive channels from which the signal can be obtained, using the same transmit or receive delays and transmit or receive apodizations in adjacent or distant elements as a single aperture By this we mean the use of waves for transmit or receive beamforming with greater intensity than transmit or receive with a single element.

The present invention includes an apparatus and method for performing arbitrary beamforming in an arbitrary Cartesian coordinate system at high speed and with high accuracy. In order to solve the problem, we target electromagnetic waves, oscillating waves (mechanical waves) including sound waves (compression waves), shear waves, surface waves, etc., or waves such as heat waves, reflected waves, transmitted waves, scattering Waves (such as forward scattered waves or backscattered waves), refracted waves, surface waves, shock waves, those waves emanating from self-emanating wave sources, waves emanating from moving bodies, or unknown wave sources Appropriately configured digital signal processing algorithms (in digital circuitry or software implementations) or analog or digital hardware are used for the observed waves, such as those coming from .

The hardware configuration may include a device equipped with an ordinary digital beamformer phasing and summing device of each wave device, and a device having an arithmetic function for performing digital wave signal processing. Alternatively, a digital circuit that implements the calculation may be constructed and used. Other necessary devices include, at a minimum, generally used transducers, transmitters, receivers, received signal storage devices, and the like, which will be described in detail later. Harmonic waves are also processed. Beamforming with virtual sources and virtual receivers is also performed. Parallel processing is also performed to generate multiple beams at the same time.

Further, in the present invention, in order to realize high-speed and high-precision processing, in addition to analog devices such as analog filtering and level adjustment by analog amplification or attenuation, analog signal processing devices (for driving signal waveforms, etc.) Efficient application of linear elements (especially nonlinear elements) for changing waveforms such as emphasizing or attenuating features, and devices, computers, and PLDs (Programmable Logic Devices) with general-purpose computing power as described above. ), FPGA (Field-Programmable Gate Array), DSP (Digital Signal Processor), GPU (Graphical Processing Unit), or microprocessors, etc. may be used, but dedicated computers, dedicated digital circuits, or Digital processing may also be performed on the stored signals using dedicated devices.

It is important that these analog devices, AD converters, memories, and devices that perform digital signal processing (multi-core, etc.) have high performance. Broadband wireless communication is also important. In particular, in the present invention, in addition to the case where those functional devices are appropriately mounted on one chip or substrate (the case where they are detachable), those functional devices are directly mounted on one chip or substrate. (including lamination) is desirable. Parallel processing is also important. If the device is non-detachable, when the computer also serves as the control unit, it is possible to obtain much higher security performance than that obtained under normal program control. On the other hand, current legislation will likely require more disclosure of what is being processed.

According to one aspect of the present invention, it is possible to use a digital beamformer with a digital operation function and perform arbitrary beamforming at a higher speed and with a higher degree of accuracy than with conventional methods. becomes. The present invention includes realizing fast and highly accurate arbitrary beamforming in arbitrary orthogonal coordinate systems, including curvilinear coordinate systems. When multiple beams or waves with different wave parameters or beamforming parameters such as ultrasonic waves are generated, even if the received echo signals received by the receiving transducer are superimposed, beamforming may be performed at once. There is also
Any beamforming, including conventional beamforming, can be implemented using the present invention . In particular , according to the present invention, the problem of requiring a long processing time when aperture elements constitute a two-dimensional or three-dimensional distribution or a multi-dimensional array can be effectively solved, and beamforming can be performed. The high speed of is even more effective. Of course, conventional DAS processing may also be used.

That is, according to the present invention, a transmitting or receiving transducer array device (sometimes used for both transmission and reception) or a sensor array device with an arbitrary aperture shape can be used to perform arbitrary beamforming at high speed and high efficiency. Precision can be realized by digital processing. Virtually any focusing, any steering (deflection), any apodization can be performed with an array device having any aperture shape.

For example, in the field of medical ultrasound imaging, in addition to Cartesian coordinate systems with linear transducers, conveyor and sector scans, or IVUS (intravascular ultrasound), physically transmit, receive, and digitize using polar coordinate systems. What you do is popular, and it is used properly depending on the observation target. For example, sector scanning is usually performed when observing the dynamics of the heart through the sternum gap. Also, in array-type aperture geometries or otherwise in PVDF (polyvinylidene fluoride)-based transducers, the apertures may be deformable. In other words, according to the present invention, it is possible to process digital signals obtained from waves transmitted and received in an arbitrary coordinate system, and obtain signals directly beamformed in an arbitrary coordinate system such as an image display system. A composite transducer of PVDF that can generate high-frequency ultrasonic waves and PZT (lead zirconate titanate) that can generate low-frequency but high-power ultrasonic waves is used as a transducer for broadband or that can generate multiple carrier frequencies at the same time. sometimes used. For generating a plurality of different ultrasonic waves, elements with different sizes, element intervals, or element thicknesses are arranged in an array, or variously configured arrays are stacked. There are cases.

In the multi-static aperture synthesis, echo data frames composed of echo signals received at the same position among a plurality of reception positions that exist with respect to a transmission position are generated by the number of reception elements , and each echo data frame is Monostatic aperture synthesis processing is performed and the processing results are superimposed . As a result, echo data can be generated by aperture plane synthesis processing the number of times equal to the number of channels of the reception aperture. Faster than methods that generate signals.

Also, based on this multi-static processing according to the present invention, dynamic reception and steering at the time of popular transmission fixed focus can be performed at high speed and with high accuracy .

Regarding the coordinate system, in the present invention, for example, in the signal processing of convex, sector scan, or IVUS, based on the echo data transmitted and received in the polar coordinate system, the Echo data can be generated directly in the Cartesian coordinate system.

Further , according to the present invention, imaging with a high SN ratio and high resolution using a virtual source can be performed at high speed. Furthermore, according to the present invention, beam frequency modulation, broadband, multi-focus, parallel processing, virtual sources, virtual receivers, etc. performed under linear processing or non-linear processing can be performed at high speed and with high precision under digital processing. can be realized. The present invention is also useful in beamforming optimization, which requires computational complexity.

In addition to the beamforming of the present invention , in various beamforming such as Fourier beamforming and conventional DAS processing, an array type aperture element group (each element is independently driven and independently obtains a received signal) 1 by applying the same transmission or reception delay and transmission or reception apodization to adjacent or distant elements and using them as one aperture. This includes using beamforming for transmission or reception with waves that are stronger than the transmission or reception with the elements. For example, in a one-dimensional array type transducer, if the element width and element spacing are shortened in order to improve the spatial resolution in the axial and lateral directions, the intensity of transmitted and received waves will weaken (the present inventor , high-precision displacement vector observation is realized at an element pitch of 0.1 mm or less). This is also the case when using a two-dimensional array or a higher-dimensional array (the element width and the element spacing in the dimensional direction are shortened). Even when the thickness of the element is reduced to increase the frequency, or when using PVDF or the like, which has a lower transmission strength than PZT or the like in ultrasonic waves, the strength of the wave becomes weak. It is effective in such a case. By the way, if the element pitch is coarse, signals with aliasing in the element array direction (originally digital space) will be received and beamforming will be performed, so the angular spectrum of the received raw signal or the signal after beamforming Filters out the signal in the selected band. If the element width is reduced and the element pitch is shortened as described above, the band becomes wide in the horizontal direction as can be confirmed by the angular spectrum, and a signal with a wide band in the horizontal direction can be generated by beamforming. need to be processed. These processes are necessary in the case of all beamforming processes. Since beamforming signals can be generated within the signal band of the angular spectrum of the received raw signal, the steering angle achievable with the element array can be ascertained.

As described above, according to the present invention, electromagnetic waves, vibration waves (mechanical waves) including sound waves (compression waves), shear waves, shock waves, surface waves, etc., or waves such as heat waves, Arbitrary beamforming, with or without transmit or receive focusing, transmit or receive steering, and transmit or receive apodization, even if the transmit and receive coordinate systems and the coordinate systems that generate the beamformed signals are different. can be performed with high accuracy and speed based on digital processing.

This not only improves the frame rate when displaying images of beamformed signals, but also provides high spatial resolution and high contrast in terms of image quality, as well as displacement and deformation using beamformed signals. Or, if the temperature or the like is measured, the measurement accuracy is also improved. High-speed processing is extremely effective in multidimensional imaging using multidimensional arrays. Although the present invention includes approximation calculations in digital operations , it is not easy to come up with because it can achieve highly accurate beamforming compared to DAS processing that includes conventional interpolation approximations with low accuracy .
In addition, according to another aspect of the invention, a new Hilbert transform using (partial) differential processing or (fast) Fourier transform enables high-speed imaging of waves, displacement (vector) measurement, temperature measurement, etc. The application becomes viable (the former is faster than the latter). When multiple beams or waves with different wave parameters or beam forming parameters such as ultrasonic waves are generated in each time phase, the number of received signals received by the receiving transducer increases and beam forming or Hilbert transform is performed. is effective in such cases because the number of times that Multiple beamformed signals may be superimposed and Hilbert transformed at once, which is also useful. When the physical aperture elements form a two-dimensional or three-dimensional distribution or a multi-dimensional array, the problem of requiring a long processing time can be effectively solved, and the high-speed Hilbert transform becomes even more effective. .

1 is a representative block diagram showing a configuration example of a measurement imaging device or communication device according to a first embodiment of the present invention; FIG. FIG. 2 is a representative block diagram showing in detail a configuration example of the apparatus body shown in FIG. 1; FIG. 4 is a schematic diagram showing an arrangement example of a plurality of transmission aperture elements in a transmission transducer; FIG. 2 is a block diagram showing the configuration of a receiving unit or receiving device equipped with a phasing adder and its peripheral devices; Schematic diagram of transmission of polarized plane waves. 4 is a flow chart showing digital signal processing during polarized plane wave transmission. FIG. 4 is a schematic diagram of a case where a wide wave is transmitted in the direction of the angle θ in the polar coordinates (r, θ) in the direction of the radius r (cylindrical wave transmission). Schematic diagram of the case of transmitting a wide wave in the direction of the angle θ in the polar coordinate system (r, θ) in the direction of the radius r (cylindrical wave transmission) using a virtual source installed behind a physical aperture with an arbitrary shape. . Schematic diagram of the position of a physical aperture with an arbitrary aperture shape, or the generation of another aperture or wave behind or in front of it. Schematic diagram of monostatic aperture synthesis. Schematic diagram of spectra generated by steering in monostatic aperture synthesis (θ is the steering angle). Schematic diagram of multi-static aperture plane synthesis. Schematic of fixed focusing using a linear array. 4 is a flowchart showing digital signal processing during cylindrical wave transmission. Schematic diagram of fixed focusing using a convex array. 4 is a flowchart showing migration processing when polarized plane waves are transmitted. FIG. 5 is a schematic diagram for explaining an example of phase aberration correction when a linear one-dimensional array transducer is used and steering is not performed; FIG. 5 is a schematic diagram for explaining an example of phase aberration correction when steering is performed using a linear one-dimensional array transducer; FIG. 4 is a schematic diagram for explaining an example of phase aberration correction when steering is performed using a two-dimensional array transducer; FIG. 10 is a schematic diagram of motion compensation performed by moving a search area provided in the next frame to be processed in the case of two-dimensional using an estimated value of a displacement vector of a point of interest or a local area containing the point of interest. 4 is a flowchart showing an example of signal processing by Fourier transform using Jacobian operation; FIG. 4 shows a numerical phantom used in the simulation; FIG. 4 is a diagram showing sound pressure waveforms of transmission pulses used in the simulation; Fig. 3 shows an image obtained using method (1) in polarized plane wave transmission; Fig. 3 shows an image obtained using method (1) in polarized plane wave transmission; FIG. 4 shows deflection angles and errors obtained using method (1) in polarized plane wave transmission versus set angles; FIG. 4 shows the deflection angle error obtained using method (1) in polarized plane wave transmission. FIG. 4 shows an image obtained using method (1) with the compound method in polarized plane wave transmission. FIG. 4 shows the point spread function realized using method (1) in polarized plane wave transmission; FIG. 11 shows an image obtained using the migration method of method (6) in polarized plane wave transmission. FIG. 10 is a diagram showing an image obtained by monostatic aperture plane synthesis of method (2); FIG. 10 is a diagram showing an image obtained by multi-static aperture plane synthesis of method (3); FIG. 10 is a diagram showing a point spread function realized by multi-static aperture synthesis of method (3); Fig. 10 shows an image obtained by fixed focusing transmission of method (4); Shows an image obtained by method (5-1) in cylindrical wave transmission using a convex array and an image obtained by method (5-1') in cylindrical wave transmission using a linear array. figure. FIG. 10 is a diagram showing an image obtained by fixed focusing transmission of convex array and method (5-2). Fig. 2 shows an example of two steering beams in the two-dimensional case and the lateral modulation of their superposition; Fig. 4 shows an example of 4 steering beams in the 3D case and the lateral modulation of their superposition; FIG. 11 is a block diagram showing a configuration example of an imaging apparatus according to a third embodiment of the present invention; The block diagram which shows the structural example of the imaging device which concerns on the 4th Embodiment of this invention, and its modification. FIG. 4 is a schematic diagram showing an arrangement example of a plurality of transducers; FIG. 4 is a diagram for explaining the form of waves when using a one-dimensional transducer array; FIG. 4 is a diagram showing the beam direction, the angle of arrival direction of waves, and the center of gravity of the spectrum in the spatial domain and the frequency domain in the case of two-dimensional measurement. FIG. 2 shows in two-dimensional space two deflected beams used for the lateral modulation method; FIG. 11 is a diagram showing an example of processing by rereading the frequency coordinate axis when aliasing occurs in the two-dimensional spectrum of the band 2A in the depth direction due to demodulation during two-dimensional horizontal modulation. FIG. 10 is a diagram showing an example of processing by rereading the frequency coordinate axis when aliasing occurs in the two-dimensional spectrum of the horizontal band 2B due to demodulation during two-dimensional horizontal modulation. FIG. 4 is a diagram showing changes in the spectrum of echo signals according to an embodiment of the present invention; FIG. 4 shows changes in the autocorrelation function of an echo signal according to one embodiment of the present invention; FIG. 4 shows changes in the autocorrelation function of an echo signal according to one embodiment of the present invention; FIG. 4 shows changes in the autocorrelation function of an echo signal according to one embodiment of the present invention; FIG. 10 is a diagram showing changes in a B-mode echo image according to one embodiment of the present invention; FIG. 10 is a diagram showing changes in a B-mode echo image according to one embodiment of the present invention; FIG. 10 is a diagram showing changes in a B-mode echo image according to one embodiment of the present invention; FIG. 2 shows images of displacement vector, strain tensor, and relative shear modulus measured in an agar phantom according to one embodiment of the present invention. FIG. 10 illustrates sound pressure variation according to one embodiment of the present invention when using a concave HIFU applicator;

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, the same code|symbol is used for the same component and the overlapping description is abbreviate|omitted. The device according to the invention can be used as a metrology imaging device as well as a communication device. In the following, when waves are sound waves such as ultrasonic waves, sound pressure or particle velocity, and when mechanical waves are compression waves (longitudinal waves) or shear waves, shock waves, surface waves, etc., stress waves or Image signals of electric field or magnetic field when targeting distorted waves and electromagnetic waves, and image signals of transmitted waves, refracted waves, reflected waves, and scattered waves (forward scattered waves, backscattered waves, etc.) of temperature or heat flux when targeting heat waves. A case of generation will be described.

<<First Embodiment>>
First, the configuration of the measurement imaging apparatus or communication apparatus according to the first embodiment of the present invention will be described. FIG. 1 is a representative block diagram showing a configuration example of a measurement imaging device or communication device according to the first embodiment of the present invention. As shown in FIG. 1, the measurement imaging device (or communication device) according to the present embodiment includes a transmitting transducer (or applicator) 10 as transmitting means, a receiving transducer (or receiving sensor) 20 as receiving means, It comprises a device body 30 , an input device 40 , an output device (or display device) 50 and an external storage device 60 .

FIG. 2 is a representative block diagram showing in detail a configuration example of the apparatus main body shown in FIG. The device body 30 mainly includes a transmission unit 31, a reception unit 32, a digital signal processing unit 33, and a control unit . Here, the receiving unit 32 may include the digital signal processing unit 33 . 1 and 2 are moderately simplified block diagrams, and the present embodiment is not limited to these, and the details of the present embodiment are as follows. As an example, communication between the above-described devices, between units in the device main body 30, and within each unit is performed appropriately based on wired technology or wireless technology, and they may be installed at remote locations. The device main body 30 is composed of a plurality of such units, and is so called for convenience.

<Transmitting Transducer>
The transmission transducer (or applicator) 10 shown in FIG. 2 generates and transmits waves by driving signals supplied from a transmission unit 31 in the device main body 30 . In this embodiment, a plurality of transmit aperture elements 10a of the transmit transducer 10 form an array.

FIG. 3 is a schematic diagram showing an arrangement example of a plurality of transmission aperture elements in a transmission transducer. 3(a1) shows a plurality of transmission aperture elements 10a densely arranged in a one-dimensional array, and FIG. 3(b1) shows a plurality of transmission aperture elements 10a sparsely arranged in a one-dimensional array. is shown. FIG. 3(a2) shows a plurality of transmission aperture elements 10a densely arranged in a two-dimensional array, and FIG. 3(b2) shows a plurality of transmission aperture elements 10a sparsely arranged in two dimensions. is shown. FIG. 3(a3) shows a plurality of transmission aperture elements 10a densely arranged in a three-dimensional array, and FIG. 3(b3) shows a plurality of transmission aperture elements 10a sparsely arranged is shown.

Each transmit aperture element 10a has a rectangular, circular, hexagonal, or other aperture shape, and may also be flat, concave, convex, and the array may be one-dimensional, two-dimensional, or , and three-dimensional ones. The directivity of one transmission aperture element 10a is determined by the frequency and bandwidth of the wave to be generated and the shape of the aperture of the transmission aperture element 10a. In order to have directivity in two directions, so-called apertures facing two orthogonal directions may be counted as one element, and so-called apertures in three orthogonal directions may be counted as one element so as to have directivity in three orthogonal directions. In some cases, the one suitable for the above is counted as one element. There are also elements with apertures oriented in more than three directions such that they are directional in more than three independent directions. They differ depending on the position and may be mixed.

The transmission aperture elements 10a may exist spatially densely or sparsely (separate positions), but this embodiment will be described without particular distinction from one-dimensional to three-dimensional array types. Aperture element arrays can be linear (flat arrangement of elements), convex (convex arc-shaped arrangement), focused (concave arc-shaped arrangement), or circular (e.g., used for IVUS in medical ultrasound, etc.). ), a spherical shape, a convex or concave spherical shell shape, and other shapes arranged in a convex or concave shape. , but not limited to these. By appropriately driving these aperture element arrays, transmission and steering of waves such as plane waves whose wavefront spreads widely in the horizontal direction, aperture synthesis, fixed transmission focus, etc. are performed, and one beam or generated A transmitted wave with a wavefront is realized.

Regarding electronic scanning, as will be described in detail later, in order to generate one transmission beam or transmission wave having a wavefront, the driving is performed by independent drive signals generated by a plurality of transmission channels provided in the transmission unit 31 shown in FIG. As many transmit aperture elements 10a as there are signals can be driven independently. A transmit aperture element array used to generate a transmit wave with one transmit beam or wavefront is also referred to as a transmit effective aperture. In addition, to distinguish from the physical aperture element array, which collectively refers to all aperture elements, the transmission aperture realized by the simultaneously driven transmission aperture elements 10a can also be referred to as a transmission sub-aperture element array, or simply as a transmission sub-aperture. be.

When the object to propagate waves (communication object) is wide or to observe the entire region of interest at once, it is possible to have as many transmission channels as the total number of aperture elements in the physical aperture element array and use all of them at all times. However, in order to reduce the cost of the device, electronically switching the transmit channel to shift the transmit sub-aperture element array (electronic scanning) or by mechanically scanning the physical aperture element array (mechanical scanning) , the waves may be transmitted over the entire region of interest using a minimal number of transmission channels. Electronic scanning and mechanical scanning may be performed together when the target to propagate waves (communication target) is wide or when the size of the observation target is large.

When sector scanning is performed, a spatially fixed aperture element array of the type described above is electronically driven and scanned (electronic scanning), or the aperture element array itself is mechanically scanned. (mechanical scanning), or both may be performed together. Classical aperture plane synthesis uses electronic scanning to electronically drive each element of an aperture element array, or mechanical scanning of one aperture element to transmit at different locations to construct a transmit aperture array. , the transmission unit 31 may have the number of transmission channels equal to the number of elements of the physical aperture array, but the number of transmission channels can be reduced by using a switching device. need. When transmitting polarized waves, the transmission unit 31 requires at least the number of channels that is obtained by multiplying the number of elements to be driven at one time by the number of polarized waves.

<Receive Transducer>
The receiving transducer (or receiving sensor) 20 shown in FIG. 2 may also serve as the transmitting transducer 10, but may be an array type sensor used separately from the transmitting transducer 10 and dedicated to receiving. Therefore, the receiving transducer 20 may be set at a different position than the transmitting transducer 10. FIG. Also, the receiving transducer 20 may be sensitive to waves different from those generated by the transmitting transducer 10 . Such receiving transducers 20 may be co-located or integral with the transmitting transducers 10 .

In the receiving transducer 20 in this embodiment, as in the transmitting transducer 10, at least one or more receiving aperture elements 20a form an array. is transmitted to the receiving unit 32 (FIG. 2). Similar to the transmit aperture element 10a, the receive aperture element 20a may have rectangular, circular, hexagonal, or other aperture shapes, and may vary from flat, concave, or convex, and the array may be one It can be dimensional, two-dimensional, or three-dimensional. The directivity of one reception aperture element 20a is determined by the frequency and bandwidth of the wave to be received and the shape of the aperture of the reception aperture element 20a. The number of apertures in one element differs depending on the position and may be mixed. Moreover, there are cases where the reception aperture elements 20a are spatially dense or sparse (separate positions), and are not distinguished from the array type here (see the example of the transmission array in FIG. 3).

Similar to that of the transmitting transducer 10, the aperture element array may be of a linear type (an array of elements is flat), a convex type (an array of convex arcuate elements), a focus type (an array of concave arcuate elements), or a circular type (for example, (used for IVUS in medical ultrasound, etc.), spherical, convex or concave spherical shells, and other shapes arranged in a convex or concave shape. Various aspects are taken and are not limited to these. Wave motion is received using these aperture element arrays, and wave reception and steering, aperture synthesis, fixed reception focus, dynamic focus, etc. of waves such as the above-mentioned plane waves whose wavefront spreads widely in the horizontal direction are performed. A received wave with a beam or generated wavefront is realized.

Transducer apertures (elements) are not spatially dense, and may exist sparsely (separate positions). Received signals may be processed in the same way when using a transducer that is not called an array type. The present invention will be described. For example, each radar may or may not form an array if there are radar apertures at remote locations on land.

In addition to radars mounted on satellites and airplanes, transducers may be used to mechanically scan the measurement target. Signals may be transmitted or received continuously densely, or spatially sparsely at distant locations. Therefore, in addition to classical aperture plane synthesis (transmission by a single aperture element), signals may be received while performing transmission beamforming. Aperture elements may exist in one dimension, or in two-dimensional or three-dimensional space. Alternatively, mechanical scanning may be performed while electronic scanning is being performed.

Regarding electronic scanning, as will be described in detail later, in order to realize one receive beam or a receive wave having a generated wavefront, receive signals that are independent of the number of receive channels provided in the receive unit 32 are received by the aperture element at once. (determines the effective receive aperture). The receive effective aperture may differ from the transmit effective aperture. To distinguish from the physical aperture element array, which collectively refers to all aperture elements, the reception aperture realized by the reception aperture element 20a used at the same time may also be referred to as a reception sub-aperture element array or simply a reception sub-aperture.

When an object (communication object) to propagate waves is wide or to observe the entire region of interest at once, the receiving unit 32 is provided with the number of reception channels equal to the total number of aperture elements provided in the physical aperture element array. To make the system cheaper, electronically switching the receive channels to shift the receive sub-aperture element array (electronic scanning) or mechanically scanning the physical aperture element array may be used, although all may be used. With (mechanical scanning) waves coming from the entire region of interest may be received using a minimal number of receive channels.

Electronic scanning and mechanical scanning may be performed together when the target to propagate waves (communication target) is wide or when the size of the observation target is large. When sector scanning is performed, a spatially fixed aperture element array of the type described above is electronically driven and scanned with alternating transmission and reception (electronic scanning), or The aperture element array itself may be mechanically scanned (mechanical scanning), or both may be performed together. Also, in classical aperture plane synthesis, transmission is performed at different positions through electronic scanning, which electronically drives each element of the aperture element array, or mechanical scanning of one aperture element, as described above. In electronic scanning, the transmission unit 31 may have as many transmission channels as the number of elements in the physical aperture array in order to construct a transmission aperture array. need.

On the other hand, regarding the reception at that time, in the monostatic type in which only the same element as the active transmission element receives signals, the reception unit 32 may be provided with a reception channel in the same manner as the transmission channel. In addition, in a multi-static type in which reception is often performed by a plurality of surrounding elements including an active transmission element, the reception unit 32 may have the number of reception channels equal to the number of elements of the physical aperture array in electronic scanning. By using switching devices, the receiving unit 32 needs to have at least as many receiving channels as the number of elements in the effective receiving aperture for both electronic scanning and mechanical scanning. When receiving polarized waves, the receiving unit 32 requires at least the number of channels obtained by multiplying the number of receiving elements by the number of polarized waves.

<Specific example of transducer>
There are various transducers 10 or 20 that can generate or receive arbitrary waves such as electromagnetic waves, light, mechanical vibrations, sound waves, or heat waves. For example, the transducer 10 may be capable of transmitting arbitrary waves to the measurement target and receiving reflected waves reflected within the measurement target, scattered waves backscattered, and the like (also serving as the transducer 20). For example, when the arbitrary wave is an ultrasonic wave, it is possible to use an ultrasonic transducer that transmits the ultrasonic wave according to the drive signal and receives the ultrasonic wave to generate a reception signal. It is well known that ultrasonic elements (PZT (Pb(lead) zirconate titanate: lead zirconate titanate), polymeric piezoelectric element, etc.) are different and the structure of the transducer is different depending on the application.

In medical applications, narrow-band ultrasound has historically been used to measure blood flow, but the inventors of the present application have found that the displacement and strain of soft tissue (in static cases), which has been put into practical use in recent years ), and the use of broadband transducers for (echo) imaging, including the measurement of shear wave propagation (velocity). The same is true for HIFU treatment, and continuous waves are sometimes used, but the inventors of the present application have also developed applicators using high-frequency or broadband devices in order to achieve high-resolution treatment. there is When using high-intensity ultrasound, the tissue is stimulated in a range that does not cause a heating effect, and a force source may be generated in the measurement object as described above, and a transducer for (echo) imaging is used. Sometimes. Heat therapy, force generation and (echo) imaging may be performed simultaneously. The same is true for other wave sources and transducers.

In the digital signal processing unit 33, the force sources can be temporally or spatially generated in multiples within the object, and by superposition of shear waves, the direction of propagation of the shear waves can be controlled, (viscosity) The anisotropy of the shear modulus and its propagation velocity can be measured. Nearly simultaneously generated shear waves are physically superimposed, so the shear waves can be separated under spectral analysis after observation of the shear waves through ultrasonic displacement measurement. In addition, when the shear waves generated by each force source are not physically overlapped, the ultrasonic signals are analyzed and observed, and the results are superimposed to determine the propagation direction, propagation velocity, and (viscosity) of the propagation direction. ) Shear elastic modulus (Patent Document 11, etc.) may be obtained, but by superimposing and analyzing the ultrasonic signals when each force source is generated, observing the superposition of shear waves, they can be obtained Sometimes it is done. The same is true when generating heat sources and observing heat waves and thermophysical properties. Various other processes are performed below.

The shape of the heat source, power source, and sound pressure can be adjusted by transmitting and receiving apodization, delay, and radiation intensity. A desired heat source, force source, and sound pressure can be realized. In some cases, the shape of the signal is observed with high sensitivity using a hydrophone, and in other cases, the shape is estimated by obtaining an autocorrelation function for the signal captured by the detector (Patent Document 11, etc.). , a linear or non-linear optimization is performed based on these processes. The direction of propagation of shear waves and thermal waves may also be optimized. In each of these cases, it is desirable to use the results of estimating mechanical properties and thermophysical properties.

For example, if a concave applicator is used, high intensity ultrasound can be focused at the focal point and is laterally broadband. However, since the sound pressure shape has a distribution that pulls a foot away from the focal position, the sound pressure shape can be made elliptical by processing (filtering, weighting, etc.) the spectrum obtained after receiving the reflected or transmitted wave. It can be processed (Patent Document 7). It may be applied that the spectral components in each direction of the wave or beam are confirmed as spectra in the same direction in the frequency domain. As a result, the quality of imaging is improved and the accuracy of displacement measurement is improved.

Wave parameters and beamforming parameters such as focus position with transmission focusing, plane wave, cylindrical wave, spherical wave without transmission focusing, deflection angle (including zero degree without deflection), with apodization or None, F-number, transmission ultrasonic frequency or transmission band, reception frequency or reception band, pulse shape, beam shape, etc. When multiple waves and beams are transmitted or received, the received signals are superimposed and When generating waves or beams with new characteristics that cannot be generated by wave generation by multiple transmission and reception or beamforming etc.), the same effect may be obtained by applying the same treatment. Note that the superimposition may be done in real time at the same time, or with respect to what was received at different times in the same phase of interest. Each of them may be subjected to receive beamforming and superimposed, or those which have not undergone receive beamforming may be superimposed and receive beamformed.

A received signal obtained from a single wave or beam, or a superposition of multiple waves or beams thereof, may be weighted and broadband in the frequency domain to perform super-resolution (high resolution). . This is so-called inverse filtering or deconvolution. A method such as that described in paragraph 0009 may also be used, and the observed wave may be subjected to the conjugate or inverse of the frequency response as an inversion of the beam properties. Conjugation of the observed wave or conjugation of the frequency response may also be applied (these are detection processes, the former yielding the squared envelope, the latter yielding the autospectrum, i.e. the autocorrelation function can get). In some cases, super-resolution is applied to beamformed (including aperture plane synthesis), and super-resolution is applied to data before reception beamforming or without beamforming at all (transmission and reception signals for aperture plane synthesis). Beamforming may be performed after performing image processing.

Further, in the displacement (vector) measurement, in order to increase the accuracy of the displacement component, the frequency in the direction of the displacement component should be increased. If high resolution is also required, it is necessary to widen the band. For example, the low frequency spectrum can be discarded and the frequency can be increased to improve the accuracy of displacement measurement. The amount of calculation can also be reduced. In the case of physically generating multiple waves or beams, or by dividing the spectrum by signal processing, multiple waves or beams may be generated. It may be done (dividing the angular spectrum before beamforming results in beamforming for each, so it is often better to divide after beamforming). Envelope detection, square-law detection, and absolute value detection are applied to the imaging. Speckle can be reduced and specular reflection can be emphasized by superimposing a plurality of waves or beams after detection.

These treatments can be performed not only when ultrasonic waves are used and in medical care, but also when electromagnetic waves are used and in various other fields. For example, it is possible to observe audible sound waves using ultrasound (that is, the Doppler effect), to observe sound waves and heat waves using electromagnetic waves and light, and to observe seismic waves using them. It is also possible to observe related physical properties (distribution) in conjunction.

There are two types of transducers: contact type and non-contact type. In some cases, a matching layer is used, and each wave is appropriately impedance-matched with respect to the object to be measured. Power, carrier frequency, bandwidth (broadband or narrowband, determines spatial resolution in the axial direction), waveform shape, element size (determines spatial resolution in the lateral direction), directivity, etc. Those designed for both performance are used (details omitted). As an ultrasonic transducer, there is also a composite one, such as one in which PZT or PVDF is laminated and which has both transmission acoustic power and broadband characteristics.

In the case of forced vibration by a drive signal, the frequency and band of the generated ultrasonic waves may be adjusted or coded by the drive signal (for reception, the signal within the band of the transducer , analog or digital filters may be used to select the band). In some cases, aperture elements having different characteristics such as frequency and sensitivity are arranged. Medical ultrasonic transducers were originally handy and easy to use, but recently, non-cable type transducers have come to be used together with handy-sized device bodies. For low frequency sounds (eg, audible sounds), there are speakers and microphones. Other wave transducers may be implemented in a similar manner, but are not limited to this.

Alternatively, the transducer 10 may be a transmitting transducer that generates arbitrary waves, and the transducer 20 may be a receiving transducer (sensor) that receives arbitrary waves. In that case, the transmitting transducer transmits arbitrary waves to the measurement object, and the sensor transmits reflected waves or backscattered scattered waves in the measurement object, or transmitted waves or waves transmitted through the measurement object. Refraction waves, forward scatter, etc. can be received.

For example, if the arbitrary wave is a heat wave, a heat source that is not intentionally generated such as sunlight, lighting, or metabolism in the body may be used, but an infrared warmer, heater, etc. Relatively steady ones, ultrasonic transducers that transmit ultrasonic waves for heating that are often controlled according to drive signals (sometimes generate a force source in the measurement target), electromagnetic wave transducers, lasers, etc. used. In addition, infrared sensors that receive heat waves and generate reception signals, pyroelectric sensors, microwave and terahertz wave detectors, temperature sensors such as optical fibers, ultrasonic transducers (temperature-dependent ultrasonic sound velocity, volume change, etc.) A nuclear magnetic resonance signal detector (detects temperature using chemical shift of nuclear magnetic resonance) can be used. For each wave, a suitable receiving transducer is used.

Optical digital cameras and mammography use CCD (charge-coupled device) technology, and may integrate an integrated circuit with the sensor body. The same technology is also applied to the ultrasonic two-dimensional array, enabling real-time three-dimensional imaging. A combination of a scintillator and a photocoupler is used to detect X-rays, and it has long been possible to observe them as waves. When capturing a high-frequency signal as a digital signal, it is effective to detect or modulate it in an analog manner as preprocessing, convert it to a low frequency, and store it in a memory or storage device (storage medium). Sometimes it is also digitally detected. These may be integrated by a chip or substrate along with transmitters and receivers.

Alternatively, the apertures may or may not form an array, for example, when the radar is located at a remote location on land. Apertures may also be mechanically scanned to obtain wide directivity. Apertures are spatially continuous and dense, spatially sparse at distant positions, evenly spaced, etc., under a certain regularity and irregularly under physical constraints It is sometimes installed in In addition, there are cases where the position is fixed with respect to an object to propagate waves (communication object) or an observation position, such as in the ocean, a building, or indoors. They may be dedicated apertures for wave transmission or reception. In addition, although each aperture may serve both functions, it is not limited to receiving only responses to waves transmitted by itself, but may also receive waves transmitted by other apertures. In medical care and observation of living organisms, ultrasonic waves generated by laser irradiation, called photoacoustic, are sometimes observed (in some cases, multiple wave transducers are integrated). According to the present invention, photoacoustics using a combination of an ultrasonic diagnostic apparatus and OCT can be used, for example, not only to distinguish between arteries and veins, but also to measure blood flow velocities in each (super-resolution can also be performed). can). It is also possible to intravenously inject a magnetic material that has an affinity for lesions as a contrast agent, apply vibrations such as ultrasonic waves to the affected area, and observe electromagnetic waves. Radio waves may also be used to communicate with various mobile objects.

Various transducers (arrays) used for passive observation such as seismic waves (seismographs), brain magnetoencephalography (SQUID arrays), electroencephalograms, electrocardiograms, neural networks (electrode arrays), radio waves (antennas), radars, etc. There is a thing, and it is sometimes used for the observation of the wave source. The propagation direction of incoming waves can be obtained based on multidimensional spectrum analysis (past achievements of the inventor of the present application). It is possible to determine the position of the wave source geometrically even when information about the propagation time cannot be obtained (usually, the position of the wave source is determined from the time when the wave was observed at multiple positions). It is possible. Waves can be observed not only as pulse waves and burst waves but also as continuous waves. Through any process, if the direction of arrival of the wave is known, various beams can be steered and focused in that direction for detailed observation. In those processes, receive beamforming is performed while changing the steering angle, always emphasizing the most likely direction, and the obtained image or imaging, spatial resolution, contrast, signal strength, etc. are observed, or Through multidimensional spectral analysis, the direction of the wave source can also be determined. Accordingly, the transducers used in the apparatus of the present invention may also be used for steering and may be equipped with mechanisms for electronic scanning, mechanical scanning, or both.

As transducers that can demonstrate the effectiveness of the present invention, relatively familiar typical transducers and some special transducers have been listed. Anything that can generate or receive arbitrary waves such as, but not limited to, electromagnetic waves, light, mechanical vibrations, sound waves, or heat waves can be used.

<Beamforming>
One or more beamforming at each aperture at the same time, or at the same phase when the state of the target to propagate waves (communication target) or the observation target is the same or substantially the same, or at another time or another time phase , or may be transmitted or received. Also, in a similar manner, one or more beamformings or transmissions or receptions may be performed in one combination of apertures. Similarly, each of the multiple combinations of apertures may also perform one or more beamforming or transmission or reception. In addition, new data may be generated through linear or non-linear calculations using them, including cases in which multiple results of beamforming or reception are obtained. There are times when the received signals to be processed are overlapped from the beginning, and there are times when the signals are overlapped.

In addition, for example, in a device such as a satellite or a radar mounted on an airplane that moves spatially, the mounted aperture may or may not form an array. In some cases, a wide directivity can be obtained by using the Alternatively, transmission and reception may be performed irregularly as necessary. There are various moving objects such as cars, ships, trains, submarines, and mobile robots. In addition, it may be something that moves regularly or randomly, such as a living thing, such as a distributed thing. In such cases, mobile communicators are used. An RFID (Radio Frequency Identification) tag, an IC card, or the like may be used.

At that time, not only is classical aperture synthesis (aperture synthesis based on transmission for each aperture element) performed, but reception beamforming may also be performed while generating transmission beamforming. In addition, mechanical scanning may be performed regularly or irregularly while electronic scanning is performed, so that waves can be appropriately propagated over a wide spatial range (communication), and a wide spatial range can be appropriately observed. sometimes Of course, by using a multi-dimensional array, it is possible to appropriately propagate waves over a wide spatial range (communication) or observe a wide spatial range (physical aperture) only by electronic scanning. not only is the

Mounted apertures may also serve as both apertures for transmitting and receiving waves, but they are not limited to receiving responses to the waves they themselves have sent, they can also receive waves sent by other arbitrary apertures. Sometimes. In some cases, a plurality of moving objects may have openings, or at the same time, or at different times or phases when the state of the object to propagate waves (communication object) or the observation object is the same or substantially the same. In other phases, one or more beamformings or transmissions or receptions may be performed at each aperture.

Also, more than one beamforming or transmission or reception may be performed in one combination of apertures in a similar manner. Further, in a similar manner, each of multiple combinations of apertures may perform one or more beamforming or transmission or reception. In addition, new data may be generated through linear or non-linear operations using beamforming or reception results, including cases in which multiple results are obtained. In the above, a combination of an aperture of a moving object and a fixed aperture may be used for an object to propagate waves (communication object) or an observation object.

Thus, in this embodiment, there are a plurality of transmission aperture elements 10a and a plurality of reception aperture elements 20a (one aperture element may serve as both the transmission aperture element 10a and the reception aperture element 20a), Beamforming is actively performed. In this active beamforming , arbitrary beamforming can be realized by digital processing at high speed. Also, virtually arbitrary focusing and arbitrary steering can be performed with a transducer array device having an arbitrary aperture shape.

Transmission is generally performed in an orthogonal coordinate system determined by the shape of the physical aperture element array (virtual sound sources will be described separately) in order to emphasize the direction of each aperture element. In order to generate a signal that directly expresses waves in a display coordinate system , reception digital beamforming is mainly performed. be. Also, virtual sources, virtual receivers, etc. may be used, and beamforming is performed in the same manner as in the case of physical aperture element arrays.

<Sending unit>
Next, the transmission unit 31 (FIG. 2) provided in the device body 30 will be described. The transmission unit 31 includes a plurality of transmission channel transmitters 31a. The number of transmission channels is the number of lines for sending different drive signals to the aperture elements used for one beamforming. For example, there are various forms of this transmission channel, as described below. The frequency, bandwidth, waveform and directivity of waves generated in each transmission aperture element 10 a are determined by the transmission aperture element 10 a and the transmission unit 31 .

When an impulse signal is applied to the transmission aperture element 10a, a wave motion determined by the shape (thickness, size and shape of the aperture) and material (a typical ultrasonic element is a single crystal) of the transmission aperture element 10a is generated. , the frequency and bandwidth of the waves generated by forcibly exciting the transmission aperture element 10a with a driving signal having a frequency, bandwidth and waveform (which may be encoded) generated in the transmission unit 31. Width, waveform and directivity are adjusted. The characteristics of the generated drive signal are set as parameters under control by the control unit 34 . The control unit 34 may recognize which transducer is used and the recommended settings may be made automatically, but settings or adjustments using the input device 40 are also possible.

Normally, when performing one beamforming, the transmission unit 31 is equipped with an analog or digital delay pattern in order to drive a plurality of aperture elements with drive signals with different delays. A delay pattern may be used that implements a transmit focus position, steering direction, etc. that can be selected using the . These patterns are programmable, and depending on the purpose, patterns to be used and patterns that can be selected may be installed through various media such as CD-ROM, floppy disk, or MO. In some cases, a program can be activated and a pattern can be interactively selected from the input device 40. In some cases, a delay (pattern) value can be directly entered, or in other cases, a file in which data is recorded can be read and set. There are cases. Particularly in the case of analog delay, the delay value to be used may be changed analogously or digitally, and the delay circuit or pattern itself may be replaced or switched.

In the device main body 30 (FIG. 2), a command signal for generating a drive signal (which may be encoded) for driving the corresponding transmission aperture element 10a is transmitted from the control unit 34 to the multi-channel transmitter 31a. be. These command signals may be generated based on a command signal for starting beamforming for one frame. When the transmission delay is digital, for example, a command signal sent to the transmitter 31a for the transmission aperture element to be driven first may be used as a trigger to apply a digital delay, and the command signal may be sent to each transmitter 31a. be. A digital circuit delay device is sometimes used for digital delay.

Further, the drive signal itself generated in the transmitter 31a for the element to be driven first may be multiplied by an analog delay and transmitted to each aperture element. In this case where an analog delay that does not require synchronization is used, at least one transmitter 31a can drive a plurality of transmit aperture elements 10a. Therefore, the transmission analog delay is provided before, after or within the transmitter 31a or within the control unit 34, while the transmission digital delay is provided within or before the transmitter 31a or within the control unit 34. There is

A pattern may be selected by switching an analog circuit or analog device, or a digital circuit or digital device, and the delay of those delay devices may be changed under the control of the control unit 34. , may be programmable through installation, input settings, etc. A delay device may also be provided within the control unit 34 . Furthermore, when the control unit 34 is composed of a computer or the like as described below, the command signal with the delay applied may be output directly from the control unit 34 under software control.

The control unit 34 and the digital delay are devices or computers with general-purpose computing power, PLDs (Programmable Logic Devices), FPGAs (Field-Programmable Gate Arrays), DSPs (Digital Signal Processors), GPUs (Graphical Processing Units), or , a microprocessor, etc., or a dedicated digital circuit, or a dedicated device. They are preferably high performance (multi-core, etc.) and may also carry analog devices, AD converters 32b, memory 32c, and/or digital signal processing units 33 that perform transmit or receive beamforming operations.

In addition, the number of times of communication between devices, communication line capacity, wiring, or broadband wireless communication are important. if it is removable). Alternatively, they may be directly mounted (including lamination) on a single chip or substrate. Parallel processing may also occur. If the device is non-detachable, when the computer also serves as the control unit 34, it is possible to obtain much higher security performance than that obtained under normal program control. On the other hand, current legislation will likely require more disclosure of what is being processed.

In some, control software and delay values are directly coded or entered, or installed. The way of applying the digital delay is not limited to these. If this digital delay is implemented in the transmission delay, unlike the analog delay, an error determined by the clock frequency for generating the digital control signal will always occur. Basically, using a high clock frequency at the cost of reducing the error. On the other hand, the analog delay can be changed in an analog manner, or it can be made digitally controllable and programmable. However, the degree of freedom is lower than that of the digital delay, and in the case of cost reduction, the delay pattern mounted as an analog circuit may be switched and used.

The transmission apodization is performed by the energy of the drive signal for the aperture element and the time change of the amplitude of the waveform (which may be encoded) that changes with time. The drive signal is adjusted based on the calibration data of the conversion efficiency (capacity) of the drive signal to waves of the aperture element. For other purposes, adjustment of the drive signal may be performed for the sole purpose of calibration. The command signal sent from the control unit 34 to the transmitter 31a may represent information on the waveform and phase of the driving signal generated by the transmitter 31a in time series, or may be recognized by the transmitter 31a and specified. It may be a coded signal capable of generating a drive signal, or it may simply issue a transmission command to the transmitter 31a that generates a predetermined drive signal for the position of each aperture element to be driven within the effective aperture. It can be.

Also, like the delay, the transmitter 31a may be programmable to produce a predetermined drive signal for the position of each aperture element to be driven within the active aperture, and may take a variety of forms. Power supplies and amplifiers are used to generate drive signals, and power supplies with different amounts of power or energy to be supplied and amplifiers with different amplification levels may be switched and used, or may be used simultaneously to generate one drive signal. or may be directly set or programmable as described above, similar to the transmit delay pattern. Delay and apodization can be implemented in the same or different forms at the same or different levels of hierarchy within a transmission unit.

The transmission channels for driving the aperture elements within the effective transmission aperture are switched through switching devices such as shift registers and multiplexers, and the effective aperture at another position is used to scan the region of interest while performing beamforming. There is Also, the delay value of the delay element may be variable, and the delay pattern (delay element group) may be switched. Further, steering may be performed in multiple directions in one effective opening, and steering may be performed in multiple directions while appropriately changing the opening position and effective opening width.

In the case of switching high voltage signals, dedicated switching devices are used. In addition, the apodization value of the apodization element may be variable in the transmission time direction or the array direction of the aperture element, and the apodization pattern (apodization element group) may be switched, and the aperture position, range direction, or steering direction may be changed. The beam shape may be adjusted depending on the Specifically, a transmit element with an apodization value of zero is not active, meaning it is off, and the apodization is also responsible for switching the active elements and can also determine the effective aperture width (apodization in the aperture element array direction). function is a rectangular window, the switch is on, otherwise it is weighted on).

In addition, regarding the delay pattern and the apodization pattern, the device body 30 may have a plurality of patterns or may be programmable. In the digital signal processing unit 33 (FIG. 2) in the device main body 30, which will be described later, attenuation and scattering (forward scattering or backscattering, etc.), transmission, reflection, refraction, or frequency dispersion of sound speed in the medium during the propagation process and spatial distribution are calculated, and the delay and intensity of the wave transmitted from each aperture, the steering direction of the beam and wavefront, the apodization pattern, etc. may be optimized.

It should be noted that classical aperture synthesis includes a monostatic type and a multistatic type performed in transmission by a single aperture element, where the active transmit aperture element 10a is, as described above, under switching or apodization. can be switched. In some cases, all transmitting elements have a transmitting channel including transmitter 31a. Aperture synthesis requires generation of waves of sufficient intensity or energy, and the transmit apodization function itself is not necessarily important. Substantially, aperture synthesis is typically performed simultaneously with receive apodization in the phasing adder, and in the present invention is often performed simultaneously with receive apodization in digital signal processing unit 33 . The representative transmission units of this embodiment have been described above, but any unit can be used as long as transmission beamforming is possible, and the description is not limited to this.

<Receiving unit and digital signal processing unit>
Next, the receiving unit 32 and the digital signal processing unit 33 (FIG. 2) provided in the device body 30 will be described. The receiving unit 32 includes a plurality of receiving channel receivers 32a, an AD converter 32b, and a memory (or storage device or medium) 32c. The frequency, bandwidth, waveform, and directivity of the received signal generated in each reception aperture element are determined by the reception aperture element 20 a and the reception unit 32 . When waves arrive at the reception aperture, a reception signal is generated that is determined by the shape (thickness, size and shape of the aperture) and material (a typical ultrasonic element is a single crystal) of the reception aperture element 20a. The frequency, bandwidth and directivity of the received signal generated by filtering in unit 32 (which may also be done by an analog amplifier) are adjusted. The generated received signal is substantially adjusted based on the filter parameters (frequency characteristics such as frequency and bandwidth) set under control by the control unit 34 . The control unit 34 may recognize which transducer is used and the recommended settings may be made automatically, but settings or adjustments using the input device 40 are also possible.

A typical digital receiving unit or digital receiving device has such a function and also has a delay-and-sum function. That is, the DAS processing in a digital receiving unit or a digital receiving apparatus is to apply phasing processing to a plurality of received signals and add the plurality of phasing-processed received signals. As the phasing process, the received signals are AD-converted in the reception channels of a plurality of reception apertures, basically stored in a memory, a storage device, or a storage medium that can be read and written at high speed, and each in the region of interest In order to phasing with respect to the position of interest, the received signal read from the storage destination is subjected to interpolation approximation processing in the spatial domain and a high-speed delay is applied. (past inventions of the present inventor, patent document 6, non-patent document 15, etc.). In addition, in the storage destination, the received signals of each reception aperture may be stored at positions according to the reception delay, and these received signals may be read out and added, or may be further subjected to the above processing. It may be added.

FIG. 4 is a block diagram showing a typical configuration of a receiving unit or receiving apparatus equipped with a phasing adder that realizes phasing addition processing, and its peripheral devices. The receiving unit (or receiving device) 35 shown in FIG. 4 includes a receiver 35a for a plurality of receiving channels, an AD converter 35b, a memory (or storage device or storage medium) 35c, and performs phasing and addition processing. A phasing adder 35d for performing digital signal processing, and a data generating unit 35e for performing digital signal processing on the generated image signal. For example, the other data generation unit 35e generates image display data, calculates displacement based on the Doppler method, calculates temperature, etc., for the object by high-order calculation. do the analysis.

Dynamic focusing is performed by performing this phasing and addition process at each position within the region of interest. Originally, dynamic focusing was a term used in the receive range direction at the effective aperture, but in fact the receive digital beamforming implemented by the present invention is not so limited. The receiving unit 32 in the embodiment of the present invention shown in FIG. Arithmetic processing including approximation calculation, and high-speed digital beamforming (Fourier beamforming, etc.) that does not require approximation processing and which is different from them, may also be performed. The receiving unit 32 may also perform phase rotation processing in the frequency domain (Patent Document 6, Non-Patent Document 15, etc.). Therefore, in the embodiment of the present invention, the digital signal processing unit 33 shown in FIG. 2 is used instead of the phasing adder 35d shown in FIG. In the digital signal processing unit 33, other data as described above may be generated based on the image signal. For example, the digital signal processing unit 33 performs beamforming and lateral modulation on the received signal generated by the receiving means such as the transducer 20 to generate the multidimensional received signal and the generated multidimensional received signal. Hilbert transform processing is performed on the multidimensional received signal.

In particular , a feature of the receiving unit 32 in the embodiment of the present invention is that, in order to realize high-speed and high-precision processing, the reception signal normally generated by the reception aperture element 20a is level-adjusted by analog amplification or attenuation. In addition to ensuring signal strength and noise reduction using analog devices such as analog filtering (which is programmable and operates under the frequency characteristics and parameters set through the control unit 34), analog signal processing is digital Utilizing the advantage of being faster than signal processing, including effectively using a device that performs linear or particularly nonlinear analog signal processing as necessary, A/D converts the signal obtained through those processes, The resulting digital signal is stored in a fast read/write memory (or storage device or medium) 32c.

In addition, as the digital signal processing unit 33 to be mounted, a device or calculator having general-purpose calculation processing capability, PLD (Programmable Logic Device), FPGA (Field-Programmable Gate Array), DSP (Digital Signal Processor), GPU (Graphical Processing Unit), or a microprocessor or the like is used, or a dedicated computer, a dedicated digital circuit, or a dedicated device is used to perform the digital wave signal processing of the present invention on the stored digital signal applied.

It is important that these analog devices, AD converter 32b, memory (or storage device or storage medium) 32c, and digital signal processing unit 33 (multi-core etc.) have high performance. communication frequency, communication line capacity, wiring, or broadband wireless communication are important. ), and also their direct mounting (including stacking) on a single chip or substrate. Parallel processing may also occur.

If the device is non-detachable, it is possible to obtain much higher security performance than security obtained under normal program control when the computer also serves as the control unit 34 . On the other hand, current legislation will likely require more disclosure of what is being processed. The digital signal processing unit 33 may also serve as a control unit 34 for sending command signals to and controlling other units.

In the receiving unit 32 embodied in the present invention, a trigger signal that causes the AD converter 32b to start sampling (AD conversion) of the received signal generated in the receiving transducer (or receiving sensor) 20 (i.e., starting AD conversion) The command signal to start capturing the digital signal to start storing the digital signal in the memory (or storage device or storage medium) 32c) is similar to the trigger signal used in a normal receiving unit. For example, any command signal generated by the control unit 34 may be used such that the transmitter 31a generates a transmit signal to the transmit aperture elements 10a to be driven, and at the multiple receive aperture elements 20a of the effective aperture: In the case of receiving waves, the command signal for the element to be driven first, the command signal for the element to be driven last, or the command signal for driving another element is used. A/D conversion may be started with a digital delay of .

These command signals may be generated based on a command signal for starting beamforming for one frame. In other words, the number of occurrences of the transmission trigger signal is counted, and the hardware or hardware that reaches a predetermined number or a number that may be a programmable parameter that is appropriately set by input from the input device 40 or the like Upon confirmation in the control program, a command signal may be generated to initiate generation of a new frame. The number, as well as other parameters, may be installed through various media such as CD-ROM, floppy disk, or MO. There are various cases, such as starting a program and interactively selecting from the input device 40, directly inputting numerical values, and setting by reading a file in which data is recorded. The number can also be determined using DIP switches or the like. If a large number of reception delay patterns is not required, the received signal multiplied by the built-in analog delay pattern may be AD-converted.

In order to realize reception dynamic focusing, a conventional high-speed reception is achieved without using a method based on the Nyquist theorem by applying a product of a complex exponential function to a signal in the frequency domain, which is a past invention of the inventor of the present application. When digital delay is applied, an error occurs that is determined by the sampling interval of AD conversion. ) had to perform its slow beamforming. On the other hand, according to the present invention, as described above, if the received signal is synchronously digitally sampled, such kind of approximation error does not occur significantly , and high-speed reception digital beamforming can be performed. . In other words, such receive digital beamforming is much faster than Fourier beamforming and the method of calculating the product of complex exponential functions in the frequency domain, which is a past invention of the present inventor.

In the present invention as well as in the ordinary case, the number of reception channels is the number of lines for sending the signals received by the reception aperture element 20a used for one beamforming to the reception unit 32. FIG. In that respect, the receiving unit 32 is as follows, and there are various types of receiving channels. That is, in order to apply different delays to the signals generated by the plurality of reception aperture elements 20a when normal beamforming is performed once, the reception unit 32 is equipped with an analog or digital delay pattern as described above. A delay pattern that realizes a reception focus position, a steering direction, and the like that can be selected by the operator using the input device 40 may be used.

These patterns are programmable, and depending on the purpose, patterns to be used and patterns that can be selected may be installed through various media such as CD-ROM, floppy disk, or MO. In some cases, a program can be activated and a pattern can be interactively selected from the input device 40. In some cases, a delay (pattern) value can be directly entered, or in other cases, a file in which data is recorded can be read and set. There are cases. In particular, in the case of analog delay, the delay value to be used may be changed analogously or digitally, and the delay circuit or pattern itself may be replaced or switched to another one. Sometimes.

If the reception delay is digital, the reception signals stored in the memory (or storage device or storage medium) 32c of each reception channel are read out and phased (added). In the apparatus of this embodiment, the digital signal processing unit 33 delays the digital received signal, the digital received signal is passed through a delay device of the digital circuit, or the AD converter 32b and the memory (or storage device or A delay can be applied to the command signal to start capturing from the control unit 34 for turning on the storage medium 32c. Therefore, the digital delay can be applied at any position after the inside of the AD converter 32b or in the control unit 34. FIG.

In the case of analog delay, the delay can be applied at any position or in the control unit 34 after the reception signal is generated in the reception aperture element 20a. When an analog delay pattern is used, at least one receiver 32a can receive the received signals of a plurality of aperture elements. Therefore, in the storage destination of the received signal, the received signal of each reception aperture may be stored at a position corresponding to the reception delay, or the reception delay may not be applied at all. Digital wave signal processing, which will be described later, may be performed in the digital signal processing unit 33 (the digital signal processing unit 33 may also be able to perform normal phasing and summing processing).

Patterns may be selected by switching analog circuits or devices, or digital circuits or devices. Also, the delays of those delay devices may be changed under the control of the control unit 34, or may be programmable through installation, input settings, or the like. Furthermore, since a delay device is provided in the control unit 34, if the control unit 34 is composed of a computer or the like as described above, the delayed command signal can be controlled under software control. It may be output directly from the unit 34.

The control unit 34 and the digital delay may be devices or computers with general-purpose computing capabilities, PLDs, FPGAs, DSPs, GPUs, microprocessors, or the like, or dedicated digital circuits or dedicated devices. They are preferably high performance (multi-core, etc.) and may also carry analog devices, AD converters 32b, memory 32c, and/or digital signal processing units 33 that perform transmit or receive beamforming operations.

In addition, the number of times of communication between devices, communication line capacity, wiring, or broadband wireless communication are important. if it is removable). Alternatively, they may be directly mounted (including lamination) on a single chip or substrate. Parallel processing may also occur. If the device is non-detachable, when the computer also serves as the control unit 34, it is possible to obtain much higher security performance than that obtained under normal program control. On the other hand, current legislation will likely require more disclosure of what is being processed. Some control software and delay values are directly coded or entered, others are installed. How to apply the digital delay is not limited to these.

In the present embodiment, a trigger signal (command signal) instructing the start of AD conversion based on the trigger signal sent from the control unit 34 (FIG. 2) of the device main body 30 is the AD signal of each reception channel. It is supplied to converter 32b. According to this command signal, AD conversion of the received analog signal of each channel and storage of the digitized signal in the memory (or storage device or storage medium) 32c are started. The transmission unit 31, the reception unit 32, and the digital signal processing unit 33, under the control of the control unit 34, change the transmission aperture position, the effective transmission aperture width, or While changing the transmission steering direction, etc., the processing from transmission to digital signal storage is repeated while changing the reception aperture position, reception effective aperture width, reception steering direction, etc. each time a wave or beam is transmitted. , each time one frame of received signals is stored, the received signal group is subjected to the digital wave signal processing method used in the present invention, that is, the digital beamforming method to generate a coherent signal.

Therefore, even if the device in the present invention is equipped with the above analog or digital delay, it is not necessarily used as a delay for beamforming, and memory, storage device, or storage medium can be saved. In some cases, it is used to delay the timing of starting the AD conversion of the received signals and their storage in the memory (or storage device or storage medium) 32c, in order to effectively utilize them and shorten the access time. The reception delay in beamforming is mainly due to the digital wave signal processing performed by the digital signal processing unit 33, and the saving and shortening of the access time are significant in the present invention. In addition, physical processing using a computer, dedicated device, etc., which is different from physical beamforming at the time of transmission (for example, software beamforming using a computer, dedicated device, etc.) In the case of classical aperture synthesis without focusing, steering, apodization, etc., which may be performed at the time or at the time of reception, the transmit delay is the same as the receive delay in digital wave signal processing. Can be hung in time.

From the above, in the present invention, the receiving unit 32 must include an independent analog or digital delay, receiver 32a, AD converter 32b, and memory (or storage device or storage medium) for each receiving channel. 32c, and if necessary, level adjustment by analog amplification or attenuation, analog filters, and other analog arithmetic devices. That is, in the apparatus of the present invention, when the reception delay is used to implement this digital delay, unless a delay for beamforming is applied, an error that depends on the clock frequency does not occur as in the analog delay.

In other words, digital delay for transmission inevitably causes an error determined by the clock frequency, so it is necessary to reduce the error by using a high clock frequency at a cost, but digital delay for reception does not require such a thing. . Using a digital delay for the reception delay does not lower the accuracy, and the degree of freedom in setting the delay pattern is high. Using an analog delay for the transmission delay improves the accuracy and allows the clock frequency to be lowered. The analog delay can be analog changeable or can be digitally controllable and programmable. However, the analog delay has a lower degree of freedom than the digital delay, and in order to reduce the cost, the delay pattern installed as an analog circuit may be switched or used by replacing it with an appropriate one. If a high degree of freedom is required for setting the transmission delay pattern, it is necessary to operate the digital delay at a high clock.

Below, the coherent signal generated by the beamforming itself of the present invention is referred to as an image signal. The reception effective aperture elements and their positions are controlled in the same manner as the transmission effective aperture elements (details will be described later). Note that this digital beamforming is not necessarily performed every time one frame of received signals is stored. It may also be done every time the number of received signals is stored, which may be the number of channels of the hardware, such as set by the hardware, or a programmable parameter (there are various input means as described above). In addition, image signals generated by partial beamforming may be combined to form an image signal of one frame.

In that case, the received signals processed at successive positions in the scanning direction may be overlapping, and when synthesizing those received signals, if simple superposition is performed (overlapping in the frequency domain). combined and inverse Fourier transformed), appropriately weighted and superimposed, or simply connected. The number of received signals stored can be confirmed in the hardware or control program by counting trigger signals (command signals arriving from the control unit 34) for capturing received signals. The command signal for starting the digital wave signal processing of one frame generated by the control unit 34 for each frame can be confirmed in the same manner, and the image signal of one frame is continuously generated appropriately.

The maximum achievable frame rate depends on the form of beamforming to be performed, and is basically determined by the wave propagation speed. Determined. Therefore, it is useful to carry out the process of partially generating the image signal by parallel processing. In addition, as described above, multi-directional aperture synthesis developed in the past by the inventor of the present application, generation of reception beams at multiple positions and reception beams in multiple directions for one transmission beam, and multi-focusing are also possible. It is useful to implement them, and in order to implement them quickly, it is useful to do parallel processing. In any case, on the basis of the above transmission and reception, one frame or a portion of received signals for beamforming may be stored, and the digital wave signal processing of the present invention, which will be described in detail later, may be performed. Also, if the image signal cannot be generated in real time, the frame rate may be reduced, or the image signal may be processed off-line.

Note that the reception apodization weights the reception signal in the reception channel of each aperture element, and may be variable with respect to the range direction. It is not impossible to make it analogue variable, but it is easy to make it digitally variable. In a normal receiving unit, phasing and addition are performed at each position, each range position, etc., and are often variable. become. On the other hand, non-variable apodization is rarely implemented, but in that case, apodization is performed when level adjustment is performed by analog amplification or attenuation of the received signal generated by the aperture element.

Although apodization has a different meaning, at least level calibration may be performed based on the calibration data of the conversion efficiency (conversion ability) of the drive signal to the wave of the aperture element, and apodization may be performed at the same time as the level calibration. sometimes be In some cases, the purpose is to process them, and in other cases, to non-linearly expand or compress the dynamic range of the waveform of the received analog signal, and other analog devices such as non-linear elements are used in each receive channel. Sometimes. Since the analog devices used, including those amplifiers and the like, are programmable, the setting method can take various forms. Like other parameters, it may be set directly using various input devices. Typically, delay and apodization are implemented in the same or different implementations in the same or different hierarchical levels within the receiving unit 32, but typically in the phasing adder. can be implemented in the digital signal processing unit 33 with a high degree of freedom.

The receive channel of each aperture element within the effective receive aperture may be switched through a switching device such as a shift register or multiplexer such that a different position of the effective aperture may be used to scan the region of interest while performing beamforming. Also, the delay value of the delay element may be variable, and the delay pattern (delay element group) may be switched. Furthermore, steering in multiple directions may be performed in one effective aperture, and steering may be performed in multiple directions while appropriately changing the aperture position and effective aperture width. Access time may be reduced by conserving equipment or storage media. It is also highly effective to store frequently used data appropriately in a small memory that is easy to read and write.

In the present invention, the savings and access time reduction are significant. The groups of apodization elements that make up the apodization pattern may also be switched. The beam shape may be adjusted depending on the aperture position, range direction, and steering direction. Specifically, a receiving element with an apodization value of zero means that it is not active but off, and the apodization also serves as a switch for the active elements and can also determine the effective aperture width (apodization in the direction of the aperture element array). function is a rectangular window, the switch is on, otherwise it is weighted on). Therefore, the apodization element is on the same level as the switch.

In addition, when the delay pattern and the apodization pattern have a plurality of patterns or are programmable, the digital signal processing unit 33 in the device main body 30 performs propagation based on the response from the transmission target and the result of beamforming. Attenuation and scattering (forward scattering, backscattering, etc.), transmission, reflection, refraction, impedance, or frequency dispersion and spatial distribution of sound velocity in the process medium are calculated, transmitted from each aperture, and received at each aperture. The delay or intensity of the waves to be applied, the steering direction of the beam or wavefront, or the apodization pattern, etc., may be optimized. For the frequency characteristics of the characteristics of these media, instead of using the instantaneous frequency and instantaneous phase of the received wave for the transmitted wave such as a pulse wave, divide the frequency response of the received wave by the frequency response of the transmitted wave at each frequency. It is effective to obtain the spatio-temporal distribution of the characteristics of the medium by subjecting it to the inverse Fourier transform. At that time, as described in paragraphs 0363 and 0371, it is effective to correct the waveform distortion generated in the observation device (device) before processing. As in the case of super-resolution using division, regularization, a Wiener filter, etc. are effective because division by a small spectral component increases noise and causes a large error. Maximum likelihood estimation is also effective, and MAP (maximum a posteriori) may be used together. In addition, they may be integrated and fused. Echoes at the same location may be acquired multiple times and processed. Also, analogous to other super-resolution, it may be useful to multiply the conjugate of the frequency response instead of dividing by the frequency response. Various super-resolutions are also described herein, including those super-resolutions. When using the above instantaneous frequency and instantaneous phase, and when performing various other imaging (for example, including reflection method and transmission method such as basic ordinary echo imaging), in observation equipment (device) It is effective to correct the generated waveform distortion before processing.
In addition, instead of the frequency response of the transmitted wave, the same transmitted wave is used as a reference material (a simple one is a medium with homogeneous propagation characteristics such as an irregular scatterer distribution, or a reflector or scattering object provided at a representative position). Others, such as the body, are more elaborate, such as frequency-dependent propagation velocity, impedance, scattering, attenuation, etc., to offset the accumulation of changes in the wave spectrum caused by the properties of the medium during propagation in a homogeneous medium. It is also effective to divide by the frequency response of the received wave obtained by applying it to a non-homogeneous material, perform maximum likelihood estimation, or multiply the conjugate of the frequency response. (which is an application of linear systems).
The above calculations may be performed over the entire region of interest at once, locally depending on the spatial inhomogeneity of the point spread function represented by the propagation properties of the wave itself, or locally (the former is easier to compute). The latter local processing has a precedent in super-resolution processing (Non-Patent Documents 35, 36), and the point spread observed locally for the reference material (irregular distribution of scatterers and scatterers at representative positions) There is also the case of division using the frequency response of the function. In other super-resolution processing, it is similarly effective to perform the entire region of interest at once or to perform it locally.

It should be noted that in classical aperture synthesis, all receive elements may have a receive channel including receiver 32a. Typically, aperture synthesis is performed simultaneously with receive apodization in the phasing adder, and in the present invention, aperture synthesis may be performed simultaneously with receive apodization in the digital signal processing unit 33 .

In addition, the ultrasonic frequency, bandwidth, code, delay pattern, apodization pattern, analog device for signal processing, effective aperture width, focus position, steering direction, and , the number of transmissions and receptions required to perform beamforming, etc. are installed in each functional device in the unit through various media such as CD-ROM, DVD, floppy disk, or MO. Sometimes.

In some cases, it is possible to start a program and interactively select them from the input device 40, directly input numerical values, select them on the operation panel of the input device 40, or download a file in which data is recorded. There are various cases such as loading and setting. It is also possible to define them using DIP switches or the like. It may also be due to replacement or switching of units. In addition, when a measurement target is selected or a transducer to be used is attached to the device, the device may automatically operate under the recommended parameters by recognizing them. Subsequent adjustments are also possible. In addition, if a functional device of an ordinary receiving unit is installed, an image image obtained by using the apparatus of the present invention and an image image obtained by processing including interpolating approximation through ordinary phasing and addition as appropriate. It is possible to compare with

<Input device>
The input device 40 is used, for example, to set various parameter values as described above. The input device 40 itself includes, but is not limited to, keyboards, mice, buttons, panel switches, touch command screens, foot switches, trackballs, and the like. General-purpose memory, USB memory, hard disk, flexible disk, CD-ROM, DVD-ROM, floppy disk, or operating system (OS) and device software can be installed from storage media such as MO, version upgrade, various parameters It may also set or update a value. The input device 40 is provided with various devices capable of reading data from a storage medium, or the input device 40 is provided with an interface, and various devices are attached and used as necessary.

The input device 40 is used not only to set parameters for various operation modes of the apparatus according to this embodiment, but also to control and switch operation modes. When the operator is a human, these input devices 40 also serve as a so-called man-machine interface, but are not necessarily controlled by a human. As described above, parameter values, data, or control signals are received from other devices through various standards and connectors, or received through wired or wireless communication (that is, communication equipment having at least a reception function), It is not limited to the above example because it is possible to obtain an effect. A dedicated or regular network may be used.

These input data are stored in a memory, storage device, or storage medium inside or outside the device, and the functional devices in the device operate with reference to the stored data. Alternatively, if the functional device in the device is equipped with a dedicated memory, data is written into it to determine or update the operation setting by software, or set by hardware, or may be changed. A calculation function within the device may operate to calculate and use optimized setting parameters based on the input data, sometimes taking into consideration the resources of the device. The mode of operation may be defined by directives. In addition, waves to be measured (types and characteristics of waves, intensity, frequency, bandwidth, or sign, etc.), objects and media to be propagated (propagation velocity, physical properties related to waves, attenuation, forward scattering, backscattering, Additional information (transmission, reflection, refraction, etc., or their frequency dispersion, etc.) may be provided to analog or digitally process the received signal as appropriate.

<Output device>
A typical example of the output device 50 is a display device. The display device not only displays the generated image signal, but also displays various results measured based on the image signal as numerical values and images. used to display. The image signal is converted (scan-converted) into an image display or moving or still image format by computational processing, sometimes using a graphics accelerator. A luminance image (gray image) or a color image may be displayed, and the meaning of luminance or color may be indicated by a scale (bar) or a logo. In addition, the results may be displayed in a bird's-eye view or as a graph, and the method of displaying the results is not limited to these.

When the result is displayed, various parameter values and patterns (names) when the operation mode is in operation may be appropriately displayed simultaneously as logos or characters together with each operation mode. Supplementary information and various types of data regarding the object to be measured, which are input from the operator or other devices, may also be displayed. In addition, the display device may be used to display a GUI (Graphical User Interface) used when setting each parameter value and pattern using the input device 40, and a touch command screen may be used. It is also used to specify an arbitrary position or an arbitrary range of an image drawn by , to enlarge the display, or to display various numerical values.

As the display device, various devices such as those using CRT, liquid crystal, or LED are used, but the display device is not limited to them, such as a dedicated three-dimensional display device or the like. In addition, the output data is not necessarily directly interpreted or interpreted by humans, and the main body of the device (computer) understands the output data based on predetermined calibration data and calculations and displays the results. (For example, understanding the composition and structure of the object to be measured from the spectrum analysis of the received signal, etc.), the output data is output to another device and interpreted by another device, and the same device (for example, a robot, etc.) Or another device may apply the output data.

A single device may receive multiple types of waves and generate an image signal for data mining, fusion, etc., or another device may be used for such processing. be. The characteristics of the generated image signal (such as intensity, frequency, bandwidth or sign) may also be analyzed. In this way, data obtained by the device according to the present embodiment may be used in other devices, and substantially a communication device having at least a transmission function may also be one of the output devices 50 . A dedicated or regular network may also be used.

<Storage device>
The generated image signals and various results (numerical values, images, etc.) measured based on the image signals are stored in a memory, storage device, or storage medium inside or outside the device, which also serves as the output device 50 . Here, they are referred to as "storage devices" to distinguish them from display devices. The external storage device 60 is also shown in FIG. 2 and the like. When the image signal is stored, operation modes, setting parameter values, supplemental information about the object input from the operator or other devices, or various data may be stored with the image signal. General-purpose or special memory, USB memory, hard disk, flexible disk, CD-R (W), DVD-R (W), video recorder, or image data capture device, etc., are used as the storage device. Devices are not limited to them. A storage device is used appropriately for its application, including the amount of data to be stored, write and read times, and the like.

Image signals and other data stored in the past may be read out from the storage device and reproduced, but the main point is that the storage device is important for storing the OS, device software, or setting parameters. is. Each functional device may also have its own storage device. Removable storage devices may also be used in other devices.

The device main body 30 reads out the image signal stored in the storage device, performs high-order digital signal processing, resynthesizes the image signal (frequency modulation by linear processing or non-linear processing, broadband, multi-focus, etc.), Image processing of image signals (super-resolution, enhancement, smoothing, separation, extraction, or CG conversion, etc.), displacement and deformation of the object, or other various temporal changes, etc. The measurement results may be output, and they may be displayed on the display device.

In addition, the stored measurement results include wave attenuation, scattering (forward scattering, backscattering, etc.), transmission, reflection, and refraction. It may be stored in a storage device and used. Optimization is performed by computational functions provided in the control unit 34 or the digital signal processing unit 33 .

<Control unit>
A control unit 34 controls the operation of the entire device. The control unit 34 is composed of various computers, dedicated digital circuits, etc., and may also serve as the digital signal processing unit 33 . Basically, the control unit 34 performs wave transmission/reception and wave digital signal processing based on various control programs and various data read from the storage device in response to various requests input through the input device 40. It controls the transmitting unit 31, the receiving unit 32 and the digital signal processing unit 33 to generate an image signal.

If the control unit 34 is composed of a dedicated digital circuit, the parameters may be variable, but in some cases only fixed operations can be realized, including the case of switching the operation. When a computer is used as the control unit 34, the degree of freedom is high, including the case of version upgrade. The basics of the control unit 34 are the number of aperture elements for transmission and reception (the number of each channel), the number of beams to be generated, the number of frames (some of them continue to operate unless they are stopped without being specified), the frame rate, etc. In order to realize the various operations described above, scanning control and image signal generation control are performed by providing the transmitting unit 31 and the receiving unit 32 with the repetition frequency, transmission/reception position information, and the like. also controls the In addition, various interfaces are provided, and various devices may be used in conjunction with each other.

The device according to the present embodiment may be used as one of devices such as a general network or sensor network, may be controlled by a network system control device, or may be a control device for controlling network devices. It may be used as a device, or it may be a controller that controls a locally configured network. There may be an interface for that purpose.

<Beam forming method>
First , a useful high-speed digital A beamforming method will be described. In digital signal processing, interim data generated in the calculation process and data to be used repeatedly may be stored in a storage device or a built-in memory as appropriate. Also, storage is used efficiently. Sometimes a small size memory is useful.

The generated image signal may be displayed as a moving image or still image on an output device 50 such as a display device, or may be stored in an external storage device 60 using a recording medium such as a hard disk. When the digital signal processing unit 33 is a computer, various languages can be used and an assembler is useful. In some cases, optimization and parallel processing are performed at compile time to achieve high-speed operations. A general-purpose device such as MatLab, various control software, or a device with a graphic interface may be used, or a special device may be used.

In the following, as an example, the beamforming method used in the apparatus of the present invention is described through the case where the waves are ultrasound. The beamforming methods that can be used in this embodiment are the following methods (1) to (7). In method (7), in addition to various beamforming methods, Disclose representative observation data.

Method (1) performs wave number matching in Fourier space in planar wave transmission and/or receive beamforming during reception, including when the direction of transmission is steered. This method does not require approximation processing. Method (1) is an invention in which wavenumber matching in the depth direction and in the lateral direction is performed by dividing a complex exponential function relating to the cosine and sine of the deflection angle and multiplying the received signal by wavenumber matching when deflection is performed. Including, the accuracy of the measurement results is improved in the same way as with classical monostatic aperture synthesis. Furthermore, as method (2), high-speed digital processing of deflection dynamic focusing by monostatic aperture plane synthesis is disclosed.

Furthermore, as method (3), high-speed digital processing of multi-static aperture plane synthesis is disclosed. In digital monostatic aperture synthesis with deflection, the center of gravity (center) or instantaneous frequency of the multidimensional spectrum of the generated image signal is ideally expressed using the deflection angle and the carrier frequency (described later). As shown in the figure, wavenumber matching is performed in the same manner as in method (1) so that the wavenumber vector has components obtained by multiplying the sine and cosine of the deflection angle by the carrier frequency. can be achieved with precision. On the other hand, in the multi-static aperture plane synthesis, echo data frames containing echo signals received at the same position among a plurality of reception positions with respect to the transmission position are generated by the number of reception elements, and in the frequency domain, each echo The data frame is subjected to the above-described monostatic digital aperture plane synthesis processing, and the superimposed processing results are subjected to inverse Fourier transform to achieve high accuracy. As a result, method (3) can generate echo data by digital aperture plane synthesis processing the number of times equal to the number of channels of the receive aperture, generate so-called low spatial resolution image signal frames and superimpose them to obtain high spatial resolution image signals. It is much faster than the conventional DAS (Delay and Summation) method of generating frames.

By the way, in the DAS method, phasing is interpolated and approximated in the spatial domain and the received signal is delayed at high speed, and in the frequency domain the delay is applied with high accuracy ( The past work of the inventors of the present application), which summates the received signals in the spatial domain. The former is fast, but the conventional method has low accuracy, and the present invention, which will be detailed later, has high accuracy (paragraph 0384). The latter is highly accurate but extremely slow.

As method (4), based on the beamforming of method (1) and method (3), a method of performing digital dynamic focus reception with high accuracy during transmission fixed focus is disclosed. As method (5), for application to convex, sector scan, or IVUS, echo data transmitted and received in the polar coordinate system is also processed through Jacobi arithmetic, and displayed with high accuracy without interpolation approximation processing. We disclose that the echo data can be generated directly in the Cartesian coordinate system of the system.

Also, as a method (6), it is disclosed that the present invention can be used for migration processing as well, with high precision and high speed processing without performing interpolation approximation processing. All beamforming processes of methods (1) to (5) can be performed based on the migration process. Finally, as method (7), these beamforming-based applications are disclosed. These demonstrate that arbitrary beamforming based on focusing and steering can be performed.

Method (1): Beamforming for plane wave transmission and/or plane wave reception (i) Echo signal (image signal) during plane wave transmission
FIG. 5 is a schematic diagram of transmission of polarized plane waves. Plane wave transmission is a method in which all elements of a linear array transducer simultaneously radiate ultrasonic waves to radiate plane waves. When the wave number is k and a plane wave with a wave vector represented by equation (0) is transmitted (x is the scanning direction, y is the depth direction, and the position of the reception effective aperture element array is the zero of the y coordinate, ), the sound pressure field at position (x,y) is given by equation (1).

Here, A(k) is the frequency spectrum of the transmitted pulse, and has the relationship of Equation (2).

この式の角スペクトルは、式（４）によって表される。

When there is a scatterer with reflection coefficient f(x, y _i ) at depth y=y _i , the echo signal from this scatterer is expressed by Equation (3).

The angular spectrum of this equation is represented by equation (4).

If the frequency response of the probe is T(k), based on the principle of the angular spectrum, the angular spectrum at the aperture plane (y = 0) of the echo signal from the depth y = y _i is expressed by Equation (5). be done.

Therefore, by adding the angular spectra from each depth, we obtain the angular spectrum of the echo signal represented by equation (6).

Therefore, by performing wavenumber matching using equations (7) and (8), the echo signal (image signal) can be expressed as equation (9) by this inverse Fourier transform.

Thinking of transmit and receive in reverse, any transmit beamforming (e.g. polarized plane waves, polarized fixed focusing beams, steering dynamic focusing based on aperture synthesis, unpolarized waves and beams, and many others) is performed on the object to be measured, a situation can be realized in which waves arriving from the object to be measured are received as plane waves with a deflection angle θ (including the case of zero degrees). This concept has not been previously disclosed. Considering this way, it is possible to receive waves at the same or different steering angles θ (zero degrees or non-zero degrees) when transmitting arbitrary beams or waves with arbitrary steering angles (zero degrees or non-zero degrees). . In addition, any wave source, any transmission effective aperture array (for example, the same as the reception effective aperture array or different from the reception effective aperture array, having an arbitrary shape and arbitrary direction aperture, or existing at another position) , or an arbitrary wave transmitted from the same physical aperture but at a different position, such as an effective aperture), receive beamforming can be performed in the coordinate system defined by the receive aperture.

It should be noted that the deflection angle .alpha. , the plane wave is steered by the deflection angle (α+θ) (the final generated transmit deflection angle is the average). This soft steering (deflection angle θ) can be considered to reinforce the physical transmission steering (deflection angle α), or to realize purely steering of plane wave transmission in software. It can also be considered that reception steering is performed by a plane wave.

Also, in method (1), the plane wave is steered at a physical deflection angle α, a soft deflection angle θ, or both deflection angles α+θ during transmission, and is dynamically focused at a steering angle φ during reception. In this case, the soft steering described in method (2) may be performed, which will be described later (the deflection angle finally generated is the average of the transmission deflection angle and the reception deflection angle). In this case, the soft steering (deflection angle θ) can be considered to reinforce the physical transmission steering (deflection angle α), or to implement the pure steering of plane wave transmission in a soft manner. However, in addition to receive dynamic focusing (including the case where the deflection angle φ is 0 degrees), it can be considered that receive steering is performed using plane waves in a software manner.

In these cases, software transmission and reception can be reversed. The processing is the same (equivalent) even if the soft polarized plane wave transmission (including the case where the deflection angle is 0 degrees) and reception deflection dynamic focusing (including the case where the deflection angle is 0 degrees) are exchanged. It is also possible to interpret a signal generated by transmission beamforming as being received by a physically polarized plane wave (including the case where the deflection angle is zero degrees) and beamformed. Normally, it is not rational to physically perform transmission dynamic focusing regardless of the presence or absence of deflection, but it can also be interpreted as physically receiving a polarized plane wave.

The method can also be used to perform arbitrary transmit beamforming (e.g. polarized plane waves, polarized fixed focusing beams, steering dynamic focusing based on aperture synthesis, unpolarized waves and beams, and many others). can. That is, the same process using plane waves for transmission can handle any wave or beam (such as the example above) generated by physical beamforming. That is, reception beamforming (dynamic focusing, etc.) can be performed regardless of the transmission. It can also be processed at once when multiple transmissions are performed. Also, in addition to the transmission deflection angle (including the case where it is zero degrees), it is possible to apply steering of plane waves or dynamic focusing (including the case where the steering angle is zero degrees) in transmission or reception in software. (The final generated deflection angle is the average of the transmit deflection angle and the receive deflection angle). Moreover, as described above, it is also possible to reverse transmission and reception, and various combinations of beamforming are possible. In some cases, beamforming is performed in each of transmission and reception, and plane wave processing is performed on both transmission and reception in terms of software. The same applies to three-dimensional beamforming using a two-dimensional array, which will be described later.

The calculation method disclosed by _J.-y. Lu et al. (Non-Patent Documents 3 and 4) is based on the above theory. , k), then perform wavenumber matching according to equation (7) and fast two-dimensional inverse Fourier transform (also described in paragraph 0354). Wavenumber matching is performed through linear interpolation approximation or approximation with the closest data itself, and requires oversampling of the received signal for accuracy. A higher order interpolation approximation may be performed, or a sinc function may be used. In the three-dimensional case, similarly, the fast three-dimensional Fourier transform and the fast three-dimensional inverse Fourier transform are used. One feature of the present invention is that wave number matching is performed without interpolation approximation. It is also characterized by the fact that it sometimes asks for a solution.

(ii) Echo Signal (Image Signal) Calculation Procedure During Transmission and/or Reception of Plane Waves According to the Present Invention A case of transmitting and/or receiving a plane wave having a deflection angle θ will be described below. In the present invention, wavenumber matching is performed in the horizontal direction by multiplying the received signal by a complex exponential function (equation (9a)) before Fourier transforming the received signal in the space (horizontal direction), and then in the depth y direction. is multiplied by the complex exponential function (equation (9b)) excluding the matching process in the horizontal direction to provide resolution in the depth y direction, and at the same time, by the complex exponential function (equation (9c)). . Of course, the deflection angle θ can be used not only at non-zero degrees but also at zero degrees. This process is not disclosed in the prior art documents.

但し、ωが角周波数（角振動数）、ｃが音速であるときに、波数ｋ＝ω／ｃである。これより、解析信号を得る。ここでは、上記の説明に合わせて、フーリエ変換として複素指数関数の核の符号が正である処理を示しているが、通常のフーリエ変換と同様に複素指数関数の核の符号を負として計算することも可能である。いずれにせよ、後に計算する逆フーリエ変換においては、必ず符号がフーリエ変換時とは逆の複素指数関数が使用される（即ち、最初と最後の変換処理は複素指数関数の核の符号が逆であればよく、最初に一般的に言う逆フーリエ変換を実施して最後に一般的に言うフーリエ変換を実施でき、順序は逆でも良い）。他の方法（２）～（７）においても、同様である。 FIG. 6 is a flow chart showing digital signal processing during polarized plane wave transmission. The calculation procedure is as follows. First, in step S11, as shown in equation (10), the received signal is Fourier transformed with respect to time t (FFT is preferred).

However, when ω is the angular frequency (angular frequency) and c is the speed of sound, the wavenumber k=ω/c. From this, an analytic signal is obtained. Here, in line with the above explanation, the processing in which the sign of the core of the complex exponential function is positive is shown as the Fourier transform. is also possible. In any case, in the inverse Fourier transform to be calculated later, a complex exponential function whose sign is opposite to that used in the Fourier transform is always used (i.e., the sign of the core of the complex exponential function is opposite in the first and last transform processes). It suffices to perform the general inverse Fourier transform first and the general Fourier transform last, and the order may be reversed). The same applies to other methods (2) to (7).

ちなみに、式（１２）の結果は、直接に計算することでも得られる。 Next, in step S12, matching processing is performed on the wavenumber _kx for deflection, and equation (10) is multiplied by equation (11). good) yields equation (12). It should be noted that the product of the fast Fourier transform result (10) and the complex exponential function (11) with respect to time t is computed using a dedicated fast Fourier transform that directly produces the computational result.

Incidentally, the result of Equation (12) can also be obtained by direct calculation.

These two Fourier transformations decompose the received signal into plane wave components. The angular spectrum formed by each plane wave at an arbitrary depth y can be obtained by multiplying the equation (13) and shifting the phase, as described above.

Furthermore, in step S14, the matching process is simultaneously performed for the wavenumber ky by multiplying the equation (14) at the same time.

In step S15, the angular spectrum for each depth y is calculated. That is, formula (16) is obtained by multiplying formula (15).

これを複数の波数ｋ成分（又は、周波数成分）を足し合わせることにより、イメージ信号が得られる。
ここで、波数ｋ（又は、周波数）と空間周波数ｋxの積分は、順序を入れ替えることができる。従って、ステップＳ１６において角スペクトルの波数ｋ成分を足し合わせ、ステップＳ１７において横方向の波数ｋ_ｘに関して逆フーリエ変換（ＩＦＦＴ）を行っても、ステップＳ１８においてイメージ信号が得られる。この場合には、１回の逆フーリエ変換で計算ができ、計算がより高速である。他の方法（１）～（６）についても同様である。偏向のための波数のマッチングは、式（１１）及び式（１４）によって行われる。周波数領域において補間近似を通じて波数マッチングする方法（非特許文献３、４）と異なり、本発明は、近似処理を行わないため、高精度に計算できる。 The sound pressure field created by each plane wave component at depth y is obtained as Equation (17) by inverse Fourier transform (IFFT) of this in the horizontal direction x.

An image signal is obtained by adding up a plurality of wave number k components (or frequency components).
Here, the order of integration of wavenumber k (or frequency) and spatial frequency kx can be exchanged. Therefore, even if the wavenumber _k components of the angular spectrum are added in step S16 and the inverse Fourier transform (IFFT) is performed on the horizontal wavenumber kx in step S17, an image signal can be obtained in step S18. In this case, calculation can be performed with one inverse Fourier transform, and the calculation is faster. The same applies to other methods (1) to (6). Wavenumber matching for deflection is performed by equations (11) and (14). Unlike the method of wavenumber matching through interpolation approximation in the frequency domain (Non-Patent Documents 3 and 4), the present invention does not perform approximation processing, so calculation can be performed with high accuracy.

である（一般的に言う逆フーリエ変換を施した結果である）。尚、偏向角度θが０度のときには、まず、横方向ｘには１次元フーリエ変換を施せば良く、高速フーリエ変換が有効である（上式の如くに計算して行く場合には、一般的に言う逆フーリエ変換を施すことになる）。そして、計算されたR'(kx,y)に対して深さ方向yにフーリエ変換すると共に波数マッチングするべく、

と１次元処理できる。式（１６'）に従って各R'(kx,ky)を計算するよりも高速である。そして、最後に、式（１６'）又は式（１６''）にx方向とy方向の２次元逆フーリエ変換を施し、イメージ信号f(x,y)が求まる。ｘ方向とｙ方向の各々に１次元逆フーリエ変換を施しても良く、高速逆フーリエ変換が有効である（上式の如くに計算した場合には、一般的に言う２次元フーリエ変換を施すことになる）。
また、逆に最後の逆フーリエ変換において波数マッチングを共に行う場合には、まず、受信信号r(x,y)に対して２次元フーリエ変換を施してR(kx,k)を得て、最後の逆フーリエ変換において式（８）を実現すべく、横方向ｘの逆フーリエ変換において式（１１）を用いると共に、深さ方向ｙの逆フーリエ変換に式（１３）と式（１４）を用いて表される式（１５）を用い、但し、２方向に同時に波数マッチングを施すため、式（１５）においてksinθの補正は必要無く、イメージ信号は、

と求まる。尚、偏向角度θが０度のときは、まず、深さ方向yに逆フーリエ変換すると共に波数マッチングするべく、

と１次元処理し、次に、横方向ｘに１次元逆フーリエ変換を施せば良く、高速逆フーリエ変換が有効である。式（１６'''）に従って各f(x,y)を計算するよりも高速である。
また、これらにおいて、波数マッチングの内、式（１１）と式（１４）の分は、代わりに、それらの複素指数関数を用いて時空間信号にて乗算するkxとkyに関する周波数変調を施して加算しても良い。この場合には、まず、横方向ｘに１次元の変換を施せば良く、高速変換が有効であり、次に式（１３）を用いた１次元処理により深さ方向yの変換と共に波数マッチングを施せば良く、高速化できる。
３次元の場合も同様であり、３次元フーリエ変換と３次元逆フーリエ変換を行えば良い。他の方法（１）～（６）についても同様である。 Physically and mathematically, wave number matching can be performed during the first Fourier transform, and wave number matching can be performed during the final inverse Fourier transform. The same applies to other methods (1) to (6). As described in paragraph 0199, in accordance with the above description, the processing in which the core of the complex exponential function is positive as the Fourier transform is shown, but the core of the complex exponential function is negative as in the ordinary Fourier transform. It is also possible to calculate In any case, the inverse Fourier transform to be calculated later always uses a complex exponential function whose sign is the opposite of that used in the first Fourier transform (that is, the first and last transform processes have the sign of the core of the complex exponential function. can be reversed, the general inverse Fourier transform can be performed first, and the general Fourier transform can be performed last, and the order can be reversed). First, when performing wave number matching in the first Fourier transform, the received signal r(x, y) (here, unlike the equation (10), the propagation time t of the wave is (13 ) and equation (14), but since wave number matching is performed simultaneously in two directions, there is no need to correct ksin θ in equation (15),

(This is the result of applying the so-called inverse Fourier transform). When the deflection angle .theta. ). Then, in order to Fourier transform the calculated R'(kx, y) in the depth direction y and wavenumber matching,

can be processed one-dimensionally. It is faster than calculating each R'(kx,ky) according to equation (16'). Finally, a two-dimensional inverse Fourier transform is applied to the equation (16') or (16'') in the x and y directions to obtain the image signal f(x,y). A one-dimensional inverse Fourier transform may be applied to each of the x direction and the y direction, and a fast inverse Fourier transform is effective. become).
Conversely, when wave number matching is also performed in the last inverse Fourier transform, first, the received signal r(x, y) is subjected to a two-dimensional Fourier transform to obtain R(kx, k), and finally In order to realize equation (8) in the inverse Fourier transform of , equation (11) is used in the inverse Fourier transform in the horizontal direction x, and equations (13) and (14) are used in the inverse Fourier transform in the depth direction y. However, since wave number matching is performed simultaneously in two directions, there is no need to correct ksin θ in Equation (15), and the image signal is given by

Asked. When the deflection angle θ is 0 degrees, first, in order to perform inverse Fourier transform in the depth direction y and wave number matching,

and one-dimensional processing, and then one-dimensional inverse Fourier transform in the horizontal direction x, and fast inverse Fourier transform is effective. It is faster than computing each f(x,y) according to equation (16''').
Also, in these, among the wave number matching, the parts of equations (11) and (14) are instead frequency-modulated with respect to kx and ky that are multiplied by the spatio-temporal signal using their complex exponential functions. You can add. In this case, it suffices to first perform one-dimensional conversion in the horizontal direction x, and high-speed conversion is effective. It's good if you apply it, and you can speed it up.
The same is true for three-dimensional data, and a three-dimensional Fourier transform and a three-dimensional inverse Fourier transform may be performed. The same applies to other methods (1) to (6).

と表される波数マッチングが以下の如くに補間近似せずに行われる。尚、２次元の場合と同様に、段落０１９２～０１９６に記載の如くに様々なビームフォーミングに応用するに当たり、波数マッチングを式（７'）と式（８'）に従って補間近似処理を行って高速に行うことはあり、その場合には、Ｆ(ｋ_ｘ',ｋ_y',ｋ_z')が３次元逆フーリエ変換される。 In addition, by using a two-dimensional aperture element array, arbitrary waves are transmitted from a wave source located in an arbitrary direction toward a measurement target, and the waves arriving from the measurement target are received as plane waves to generate three-dimensional waves. When performing digital signal processing, for example, a Cartesian Cartesian coordinate system (x, y, z ), when the zero or non-zero deflection angle formed by the direction of reception as a plane wave and the axial direction is expressed using the elevation angle θ and the azimuth angle φ, the Fourier transform is obtained in the depth y direction and the lateral directions x and z is performed, the three-dimensional Fourier transform R′(k _x ,k,k _z ) of the received signal is given by

is performed without interpolation approximation as follows. As in the two-dimensional case, when applying various beamforming as described in paragraphs 0192 to 0196, wave number matching is performed by interpolation approximation according to equations (7′) and (8′) to perform high-speed , in which case F(k _x ', k _y ', k _z ') is subjected to a three-dimensional inverse Fourier transform.

さらに、その積を、軸方向ｙに関して分解能を持たせるべく横方向ｘ及びｚに関してフーリエ変換して得られた角スペクトルに、横方向ｘ及びｚに行われた波数マッチングの効果を除いて実施できる複素指数関数（Ｃ２２）を掛けると同時に、複素指数関数（Ｃ２３）を掛けることにより軸方向に関する波数マッチングを行い、ここで、横方向の波数がｋ_ｘ及びｋ_zで表され、

補間近似処理を行うことなく波数マッチングを行うことによって、直接的にデカルト座標系においてイメージ信号を生成することができる。即ち、各平面波成分が深さｙに作る音圧場は、これを横方向ｘとzに関する２次元逆フーリエ変換（ＩＦＦＴ）を行い、複数の波数ｋ成分（又は、周波数成分）を足し合わせることにより、イメージ信号を得る。無論、偏向角度が零度（即ち、仰角θ及び方位角φが零度）やいずれかの少なくとも１つの角度が零度のときにも使用できる。 When interpolation approximation processing is not performed in wave number matching, first, the received signal obtained by Fourier transforming in the axial direction y is multiplied by a complex exponential function (C21) expressed using wave number k and imaginary unit i. perform wavenumber matching in the lateral directions x and z;

Further, the product can be performed on the resulting angular spectrum by Fourier transforming it in the transverse directions x and z to provide resolution in the axial direction y, excluding the effects of wavenumber matching performed in the transverse directions x and z. Axial wavenumber matching is performed by multiplying by a complex exponential function (C22) and simultaneously by a complex exponential function (C23), where the transverse wavenumbers are represented by _kx and _kz ,

Image signals can be generated directly in the Cartesian coordinate system by wavenumber matching without interpolation approximation. That is, the sound pressure field created by each plane wave component at depth y is obtained by performing a two-dimensional inverse Fourier transform (IFFT) on this with respect to the horizontal directions x and z, and summing up a plurality of wave number k components (or frequency components). to obtain an image signal. Of course, it can also be used when the deflection angle is zero degrees (that is, the elevation angle .theta. and the azimuth angle .phi.

In the above calculation, only signal components within a band determined by the transmission signal or a band determined by considering the SN ratio of the received signal are subject to calculation. For example, when generating an analytic signal based on equation (10), only signals within a necessary band may be generated and stored (corresponding to downsampling). In the method or apparatus of the present invention, interpolation approximation processing is not performed when performing wavenumber matching, but by oversampling echo signals in the depth direction and lateral direction, the effect of becoming robust against mixed noise is obtained. be. The same applies to other methods (1) to (6).

In addition, by setting the position coordinates and range in the depth y direction to be calculated, the interval of the coordinates, etc., in formulas (13) to (15), formula (C22), and formula (C23), an arbitrary depth position and Image signals with arbitrary depth ranges, arbitrary depth intervals, and arbitrary densities can be generated without interpolation approximation (with or without downsampling as described in the previous paragraph). can be up-sampled, and the down-sampling is effective as long as the Nyquist theorem is satisfied.However, high-frequency signal components may be intentionally processed out of band (filtered out).Also, as described in the previous paragraph Downsampling is possible as long as the Nyquist theorem is satisfied, regardless of the presence or absence of downsampling of . In addition, in the horizontal inverse Fourier transform such as Equation (17), by determining the position coordinates and range in the horizontal x direction to be calculated (if necessary, the complex exponential function, which is a past invention of the inventor of the present application, analog-like spatial shifting by phase rotation using multiplication), image signals at arbitrary horizontal positions and arbitrary horizontal ranges can be generated without performing interpolation approximation processing, and the same inverse Fourier processing can be used. In the transformation, horizontal band narrowing (spatial densification) by excluding the frequency coordinates of the high frequency band, or horizontal band widening by zero stuffing of the angular spectrum (spatial densification) can generate image signals with arbitrary horizontal intervals and arbitrary densities without interpolation approximation.

In this manner, image signals can be generated at any desired location, extent, spacing, or density. That is, it is also possible to generate image signals at intervals shorter than the sampling interval of the received signal and at intervals shorter than the interval of the reception aperture elements. Also, the image signals can be coarsely spaced in each direction (provided that the Nyquist theorem is satisfied). In addition, as in the two-dimensional case (paragraph 0205), physically and mathematically, wavenumber matching can be performed during the first Fourier transform, and wavenumber matching can be performed during the final inverse Fourier transform. It is possible. In the case of no steering, similarly, in the conversion process involving wave number matching, the conversion process is first performed in the horizontal direction x or z direction, and finally the depth direction y can be performed, thus speeding up. . In these, among the wave number matching, the parts of the equations (C21) and (C23) may be added after performing frequency modulation by multiplying by the spatio-temporal signal using their complex exponential functions instead. good. In this case, first, one-dimensional transformation in the horizontal x or z direction is effective, and high-speed transformation is effective. It is sufficient if matching is performed, and the speed can be increased. In addition, in order to improve accuracy when wave number matching is performed through interpolation approximation according to equations (7) and (8), or equations (7′) and (8′), at the expense of an increase in the amount of calculation, It should be processed under proper oversampling. In that case, it should be noted that the number of data for Fourier transform increases, unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed. The same applies to other methods (1) to (6).

When generating a wide wave (cylindrical wave) in the angle θ direction by transmitting or receiving in the radial r direction of the polar coordinates in the convex transducer, sector scan, or IVUS (Fig. 7), or in other aperture shapes, When performing the same beamforming (cylindrical wave) using a virtual source installed behind (see FIGS. 8A (a) to (c), Patent Document 7, Non-Patent Document 8, etc.), in the above method, Cartesian coordinates (x, y) can be converted to polar coordinates (r, θ) for processing, and image signals can be generated in polar coordinates (r, θ). The same is true for spherical waves in a spherical coordinate system. In addition, as shown in FIGS. 8B (d) to (f), using a physical aperture element array represented by the polar coordinate system as described above or a physical aperture with an arbitrary aperture shape, transmission or reception, Alternatively, both plane waves may be generated and beamforming may be performed in the same manner. This is equivalent to generating a virtual linear aperture array (or plane wave) at that distance position, and if the distance position is set to zero, it corresponds to the use of a virtual linear aperture array. Distance positions can be set behind the physical aperture as well as in front, and a virtual linear aperture array (or plane wave) can be generated at those positions. A virtual linear aperture array may be used as a virtual receiver instead of a virtual source, or may also serve as a virtual source.

FIG. 7 is a schematic diagram of a case of transmitting or receiving a wide wave in the direction of the angle θ in the polar coordinates (r, θ) in the direction of the radius r (transmission of cylindrical wave). FIG. 7(a) shows cylindrical wave transmission using a convex aperture element array, FIG. 7(b) shows cylindrical wave transmission using a sector aperture element array, and FIG. (c) shows cylindrical wave transmission using an IVUS (circular) aperture element array. It should be noted that sector scanning may be performed using a flat aperture, which is shown in FIG. 7(b) with an arcuate aperture. These apertures may also be used to generate focused beams.

FIG. 8A shows the case where a wide wave is transmitted in the direction of the angle θ in the polar coordinate system (r, θ) using a virtual source installed behind a physical aperture of arbitrary shape (cylindrical wave transmission ) is a schematic diagram. FIG. 8A(a) shows cylindrical wave transmission using a linear aperture element array, and FIG. 8A(b) shows cylindrical wave transmission using a convex aperture element array. (c) shows cylindrical wave transmission using another arbitrary aperture element array. Receiving may be done in a similar way. Also, FIGS. 8B(d) to (f) show a case where a transmission plane wave is generated at an arbitrary distance position using a physical aperture element array represented by a polar coordinate system or a physical aperture having an arbitrary aperture shape. (The figure shows the case where a convex aperture element array is physically used). Assuming that the distance position is zero, it virtually corresponds to the case of using a linear aperture array (FIG. 8B(d)). The distance position can be set not only behind the physical aperture (FIG. 8B (e)) but also in front (FIG. 8B (f)), and a virtual linear aperture array (or plane wave) is generated at those positions. You can also Receiving may be done in a similar way. FIG. 8B (g) is a special case, for example, in the case of physically using a linear array type transducer, applying the case of generating a cylindrical wave using a virtual source behind the physical aperture, at an arbitrary distance position, FIG. 10 is a schematic diagram of generating a plane wave or a virtual linear array type transducer that spreads in the lateral direction; Receiving may be done in a similar way. A virtual linear array transducer may be used as a virtual receiver rather than a virtual source, or may also serve as a virtual source.

In the case of transmission focusing, there is a report in Non-Patent Document 6, and results are similarly obtained in the polar coordinates (r, θ). For example, there is an effect that a wide FOV can be obtained. As a method different from Non-Patent Document 6, as one feature of the present invention, this method (1) is used for these, and furthermore, a steering angle having a steering angle is also used at the coordinate positions of these polar coordinate systems (r, θ). It is possible to generate beams, as well as using methods (2)-(4) and (6) below, in which Cartesian coordinates (x, y) are converted to polar coordinates (r, θ) should be read and processed. However, when such beamforming is performed, it is necessary to perform interpolation processing in order to obtain the signal value at the coordinate position of the Cartesian coordinate system of the display system. Strict interpolation processing is performed over a long period of time, or interpolation approximation processing is performed as short-time processing with approximation errors. The same applies to spherical coordinates.

In addition, in these cases of beamforming in a polar coordinate system, displacement measurements can also be made, for example displacement components in the direction of radius r or angle θ can be measured, or displacement components in both directions can be measured. A displacement vector can be measured. However, in order to obtain the measurement result at the coordinate position of the Cartesian coordinate system of the display system after measurement, it is necessary to perform interpolation processing. Strict interpolation processing using the method is performed over a long period of time, or interpolation approximation processing is performed as short-time processing with approximation errors. The same applies to the spherical coordinate system.

From the results of displacement measurement, strain, strain rate (tensor), velocity (vector), and acceleration (vector) are obtained by partial differential processing using a differential filter, and furthermore, mechanical properties (such as bulk modulus and The shear modulus (for example, Non-Patent Document 7), other elastic modulus tensor of an anisotropic medium, etc.), temperature, etc. may be obtained through calculation. When performing interpolation approximation, it is often possible to shorten the calculation time by performing approximation processing and performing these calculations in the Cartesian coordinate system. It is better to display it as In other words, the only error that occurs in the process after displacement measurement is interpolation approximation when obtaining the final display data (there are cases where a plurality of display data are obtained from the same displacement data).

It should be noted that, as described above, the echo signals can also be represented in a Cartesian coordinate system and the displacement measurement and series of measurements can be performed through the interpolation process. If approximation processing is performed in interpolation processing, an error will occur, but the amount of calculation required for the whole process is small. Even when performing measurement based on processing of other echo signals, the processing can be performed as described above. The same applies to three-dimensional beamforming using a two-dimensional array.

In addition, regarding any of the above, similar processing can be performed for any orthogonal coordinate system other than the polar coordinate system.

On the other hand, similarly, when transmitting or receiving (cylindrical waves) a wide wave in the direction of the angle θ in the polar coordinates (r, θ) in the direction of the radius r in a convex transducer, sector scan, IVUS, etc. (Fig. 7) Then, using a virtual source installed behind a physical aperture with an arbitrary aperture shape, a wide wave in the direction of angle θ in the same polar coordinate system (r, θ) as in FIG. 7 is transmitted in the direction of radius r (cylindrical wave) (see FIGS. 8A(a)-(c)), the methods for generating image signals directly in Cartesian coordinates are method (5) and method (5-1), (5-1′), respectively. ) etc. Further, using a physical aperture element array represented by the polar coordinate system as described above or a physical aperture having an arbitrary aperture shape, plane waves for transmission or reception, or both are generated at an arbitrary distance position, and similarly beams are generated. Forming may also be performed, and an image signal may be generated in the Cartesian coordinate system (see FIGS. 8B(d)-(f)). This is equivalent to generating a virtual linear aperture array (or plane wave) at that distance position, and if the distance position is set to zero, it corresponds to the use of a virtual linear aperture array. Distance positions can also be set in front of the physical aperture in addition to behind it, and virtual linear aperture arrays (or plane waves) can also be generated at those positions. A virtual linear aperture array may be used as a virtual receiver instead of a virtual source, or may also serve as a virtual source. Similarly, beamforming methods are described as method (5) and methods (5-1), (5-1′), and so on. In these cases, echo signal imaging, displacement measurement, etc. can be performed consistently in the same Cartesian coordinate system. The same applies to the polar coordinate system. In such a case, it is equally possible to convert the echo signals and measurements from the Cartesian coordinate system to the polar coordinate system through interpolation processing. The same applies to three-dimensional beamforming using a two-dimensional array. Transmit focusing may also be performed.

Further, in any of the above, similar processing can be performed for any orthogonal coordinate system other than the polar coordinate system. In addition, as described above, the virtual source and the virtual receiver are not necessarily installed behind the physical opening, but may be installed in front of the opening. (Patent Document 7, Non-Patent Document 8). Thus, the present invention is not limited to the above. Also, in the wavenumber matching in these beamformings, an interpolation approximation process is sometimes performed to obtain an approximate solution at high speed. Also, any of them may be processed in the same way in the present methods (1) to (7).

In addition, in methods (1) to (7) of the present invention, the apodization processing for transmission, reception, or both of the received signals can be performed at various timings because of linear processing. That is, it can be performed by hardware at the time of reception, or by software at various timings after reception. As described above, physical apodization may occur during transmission (the same shall apply hereinafter).

As a matter of course, when beamforming is performed by receiving a transmitted wave instead of an echo signal, the coordinate y is half the round-trip distance (expressed as ct/2 using the propagation time t). is the distance (ct) from the aperture element in the coordinate system determined by the receive aperture element array.

Next, a case of performing aperture plane synthesis will be described. Aperture synthesis includes monostatic and multistatic types.
Method (2): Monostatic Aperture Synthesis FIG. 9 is a schematic diagram of monostatic aperture synthesis. Monostatic aperture synthesis radiates ultrasonic waves from one element of the array and receives echoes at that element itself. Also in aperture plane synthesis, an echo signal (image signal) can be calculated by performing wave number matching according to the procedure of FIG.

In monostatic aperture synthesis, since transmission and reception are performed by the same element, the propagation path of sound to the scatterer during transmission and the propagation path of the reflected sound from the scatterer during reception are the same. Therefore, in a Cartesian orthogonal coordinate system in which the position of the reception effective aperture element array is zero in the axial direction y coordinate, when steering is not performed (θ is zero degrees), the wave number k is doubled as shown in equation (18a). (in the case of reflected waves, s=2, and so on), and matching of the wavenumbers represented by equations (7) and (8) is performed. In the case of a transmitted wave, the wavenumber k is used instead of the wavenumber 2k (s=1, and so on).

When the steering angle θ is non-zero, the wavenumber vector (0, k ₀ ) having the wavenumber k ₀ (=ω ₀ /c) expressed using the carrier frequency ω ₀ of the ultrasonic signal is In order to generate an image signal having the vector (sk ₀ sin θ, sk ₀ cos θ) as the centroid (center) of the multi-dimensional spectrum or the instantaneous frequency, beamforming is performed to shift the spectrum (see FIG. 10). That is, wave number matching represented by Equation (18b) is performed in Equations (7) and (8).

The signal processing is performed in the same manner as in method (1), and in particular, the wave number matching is performed by applying ultrasonic wave Multiplying by a complex exponential function (equation (19a)) expressed using the carrier frequency ω ₀ of the signal is first performed in the horizontal direction, and with respect to the depth y direction, in order to have a resolution in the depth y direction, Instead of the complex exponential function (equation (9b)), the complex exponential function (equation (19a)) is multiplied by the complex exponential function (equation (19b)) excluding the horizontal matching process (equation (19a)), and at the same time, the complex exponential function (equation (9c)) Instead, it is multiplied by a complex exponential function (equation (19c)). Of course, it can also be used when the deflection angle is zero degrees. This process is not disclosed in the prior art documents.

Further, for example, in the echo method (reflection method), the _{deflection angles} of the transmission beam and the reception beam may differ. , in equations (7) and (8), under s=2, perform wave number matching represented by equation (18c).

Signal processing is performed in the same manner as in the above case where the deflection angles of the transmission beam and the reception beam are equal. In particular, the wave number matching is performed using a complex exponential function (formula (19a)) instead of (19a)) by a complex exponential function (equation (19d)) expressed using the carrier frequency _ω of the ultrasonic signal, first in the lateral direction, and in the depth y direction, the depth In order to provide resolution in the y direction, the complex exponential function (equation (19b)) is multiplied by the complex exponential function (equation (19d)) excluding the horizontal matching process (equation (19d)). The complex exponential function (formula (19f)) is multiplied instead of the exponential function (formula (19c)). Of course, it can also be used when the deflection angles θ _t and θ _r are zero degrees. This process is also not disclosed in the prior art documents.

From the equations (19a) to (19c) and the equations (19d) to (19f), the two-dimensional Fourier transform R'(k _x ,k) of the received signal is given by the equations (7) and (8) , (18b) and (18c) without interpolation approximation. On the other hand, interpolation approximation processing may be performed to perform beamforming at high speed, in which case F(k _x ', k _y ') is subjected to a two-dimensional inverse Fourier transform. In addition, as in method (1), regarding equations (18a) to (18c), it is possible physically and mathematically to perform wave number matching at the time of the first Fourier transform, and finally inverse Fourier transform. Wavenumber matching can sometimes be performed (paragraphs 0205 and 0210). However, in these cases, even during steering, in transform processing involving wave number matching, first one-dimensional transform processing is performed in the x direction, and then one-dimensional transform processing in the y direction (wave number mapping in the x direction). (19b) and (19e)) can be implemented and the processing speed can be increased. Also, in the wave number matching, the parts of Eqs. (19a) and (19c), Eqs. (19d) and (19f) are replaced by kx and Frequency modulation on ky may be implemented. In addition, these approximation processes, including the approximate wave number matching process corresponding to Eq. (18a) when the deflection angle is set to zero degrees in Eqs. (18b) and (18c), are not disclosed in the prior art document. do not have.

Further, when three-dimensional wave digital signal processing is performed using a two-dimensional aperture element array, for example, the axial direction y determined by the direction of the aperture of the flat reception aperture element array (y of the position of the reception effective aperture element array the zero or non-zero degree deflection of the generated beam direction and axis in a Cartesian Cartesian coordinate system (x, y, z) using transverse x and z coordinates) and transverse x and z coordinates. In cases where angles are expressed in terms of elevation angle θ and azimuth angle φ, the Fourier transform is a three-dimensional Fourier transform with respect to depth y and lateral x and z (wavenumbers in axial y and lateral x and z are Consider wavenumber domains ( _kx , _ky , _kz ) with k _y , k _x , and k _z , respectively), and are processed as follows in the same manner as the two-dimensional wave digital signal processing described above. be.

さらに、その積を、軸方向ｙに関して分解能を持たせるべく横方向ｘ及びｚに関してフーリエ変換して得られる角スペクトルに、横方向ｘ及びｚに行われた波数マッチングの効果を除いて実施できる複素指数関数（Ｃ４２）を掛けると同時に、複素指数関数（Ｃ４３）を掛けることにより軸方向に関する波数マッチングを行い、

補間近似処理を行うことなく波数マッチングを行うことによって、直接的にデカルト座標系においてイメージ信号を生成することができる。即ち、各平面波成分が深さｙに作る音圧場は、これを横方向ｘとzに関する２次元逆フーリエ変換を行い、複数の周波数ｋ成分を足し合わせることにより、イメージ信号を得る。無論、偏向角度が零度（即ち、仰角θ及び方位角φが零度）やいずれかの少なくとも１つの角度が零度のときも計算できる。 First, for a wave vector (0, k ₀ , 0) having a wave number k ₀ (=ω ₀ /c) expressed using a wave carrier frequency ω ₀ , a wave vector (sk ₀ sin θ cos φ, sk ₀ cos θ, The received signal was Fourier-transformed with respect to the axial direction y to perform dynamic focusing of transmission and reception with spectral shifting to generate an image signal with the centroid (center) or instantaneous frequency of the multidimensional spectrum at sk ₀ sin θ sin φ). a parameter s whose value is 2 when the y-coordinate of the transmit aperture element is _zero and whose value is 1 when the y-coordinate of the transmit aperture element is non-zero; , by a complex exponential function (C41) expressed with the imaginary unit i, performing wavenumber matching in the lateral directions x and z,

Further, the product is Fourier-transformed in the lateral directions x and z to provide resolution in the axial direction y, and the resulting angular spectrum is converted to a complex spectrum that can be implemented without the effects of wavenumber matching performed in the lateral directions x and z. Wave number matching in the axial direction is performed by multiplying the exponential function (C42) and the complex exponential function (C43) at the same time,

Image signals can be generated directly in the Cartesian coordinate system by wavenumber matching without interpolation approximation. That is, the sound pressure field created by each plane wave component at depth y is subjected to a two-dimensional inverse Fourier transform in the horizontal directions x and z, and an image signal is obtained by adding a plurality of frequency k components. Of course, the calculation can also be performed when the deflection angle is zero degrees (that is, the elevation angle θ and azimuth angle φ are zero degrees) or at least one of the angles is zero degrees.

As a result, the following wavenumber matching (formula (C44)) is performed as in formulas (7') and (8') for the three-dimensional Fourier transform R'(k _x ,k,k _z ) of the received signal. situation can be realized. Also, as in the two-dimensional case, regarding Eq. (C44), both physically and mathematically, wavenumber matching can be performed during the first Fourier transform, and wavenumber matching can be performed during the final inverse Fourier transform. It is also possible (paragraph 0229). Even during steering, in transform processing involving wave number matching, first one-dimensional transform processing is performed in the x direction or z direction, and finally one-dimensional transform processing in the y direction (after wave number mapping in the x and z directions). (C42) and (C43)) can be implemented and the speed can be increased. Also, among the wave number matching, the parts of the equations (C41) and (C43) may instead perform frequency modulation by multiplying by the spatio-temporal signal using their complex exponential functions (each, kx and kz, ky frequency modulation). In the above, this wave number matching is performed without interpolation approximation, but interpolation approximation processing may be performed according to _equation ( _C44 ) at high speed. ',k _z ') is subjected to a three-dimensional inverse Fourier transform. This process is also not disclosed in the prior art documents.

尚、送信と受信の両ビームフォーミングがソフト的に行われる開口面合成において、それらの送信と受信を入れ替えてビームフォーミングしても同処理である。 Further, for example, in the echo method (reflection method), the _{deflection angles} of the transmission beam and the reception beam may differ. ) and (θ _r , φ _r ), the signal processing is performed under s=2 in the same manner as in the above case where the deflection angles of the transmit and receive beams are equal, and in particular, Wave number matching is a complex exponential function expressed using the carrier frequency ω ₀ of the ultrasonic signal instead of the complex exponential function (Equation (C41)) in the stage prior to spatially (horizontally) Fourier transforming the received signal ( Equation (D41)) is first performed in the horizontal direction, and for the depth y direction, horizontal matching is performed instead of the complex exponential function (Equation (C42)) in order to have resolution in the depth y direction. The complex exponential function (formula (D42)) excluding the processing (formula (D41)) is multiplied, and the complex exponential function (formula (C43)) is replaced by the complex exponential function (formula (D43)). . Of course, it can also be used when the deflection angles of the transmit and receive beams are zero degrees (ie, θ _t , φ _t , θ _r , and φ _r are zero degrees). This process is also not disclosed in the prior art documents.

In aperture synthesis in which beamforming for both transmission and reception is performed by software, the same processing is performed even if the beamforming is performed by switching the transmission and reception.

As a result, the following wavenumber matching (formula (D44)) is performed as in formulas (7') and (8') for the three-dimensional Fourier transform R'(k _x ,k,k _z ) of the received signal. situation can be realized. Similarly, regarding equation (D44), physically and mathematically, wavenumber matching can be performed during the first Fourier transform, and wavenumber matching can be performed during the final inverse Fourier transform. (Paragraph 0232). Even during steering, in transform processing involving wave number matching, first one-dimensional transform processing is performed in the x direction or z direction, and finally one-dimensional transform processing in the y direction (after wave number mapping in the x and z directions). (D42) and (D43)) can be implemented and the speed can be increased. Also, among the wave number matching, the parts of the equations (D41) and (D43) may instead perform frequency modulation by multiplying by the spatio-temporal signal using their complex exponential functions (each, kx and kz, ky frequency modulation). In the above, this wave number matching is performed without interpolation approximation, but interpolation approximation processing may be performed according to the formula (D44) at high speed. In that case, F(k _x ', _ky ',k _z ') is subjected to a three-dimensional inverse Fourier transform. This process is also not disclosed in the prior art documents.

Aperture plane synthesis can generate arbitrary beamforming using echo signals collected for aperture plane synthesis (not only in monostatic method (2), but also in multistatic method (3)). By the way, the image signal can be generated by performing the processing described in methods (1) and (4) to (7) on those data). In the plane wave processing of method (1) as well, the aperture plane synthesis processing can be performed by using the code. In other words, a signal for aperture plane synthesis can be obtained by decoding the coded received signal at the time of plane wave transmission.
Steering can also be performed in dynamic focusing, as described in method (1). In method (1), physical steering is performed at the deflection angle α (including the case of zero degrees) when transmitting the plane wave, and the steering of the deflection angle θ (including the case of zero degrees) in method (1) is performed. can be interpreted as steering the plane wave at a steering angle (α+θ) (the final generated steering angle is the average). Therefore, in method (1), if a plane wave is steered at a deflection angle α, θ, or α+θ during transmission and dynamically focused at a steering angle φ (including zero degrees) during reception, The reception steering described in method (2) may be performed, and the finally generated steering angle is the average of the transmission deflection angle and the reception deflection angle. As described in method (1), this soft plane wave steering (deflection angle θ) reinforces the physical transmission steering (deflection angle α), or purely replaces the pure plane wave transmission steering in software. Alternatively, it can be considered that receiving dynamic focusing (including the case where the deflection angle φ is 0 degrees) and receiving steering using a plane wave are performed softly.

That is, in the two-dimensional case, (F41), (F42), and (F43) are combinations of (9a) and (19a), (9b) and (19b), and (9c) and (19c) respectively. can be used to do the same.

In the case of three dimensions, that is, when a plane wave is physically steered at the deflection angles (α, β) of the elevation angle α and the azimuth angle β, or when at least one angle is zero degrees , when the plane wave is steered at the deflection angles (θ ₁ , φ ₁ ) and the steering dynamic focusing is performed at the deflection angles (θ ₂ , φ ₂ ) (at least one angle is zero degree (including cases) use (G41), (G42), and (G43), which are combinations of (C21) and (C41), (C22) and (C42), and (C23) and (C43) respectively. , and the average deflection angle of the transmission deflection angle and the reception deflection angle can be finally generated.

As mentioned in method (1), the processing is the same even if the software-based transmission and reception beamforming are exchanged, and is equivalent to the exchanged beamforming. In other words, even in these cases, soft transmit and receive beamforming can be considered inversely, and any physical transmit beamforming (e.g., polarized plane wave, polarized fixed focused beam, aperture plane In software, various combinations of beamforming are possible in synthesis-based steering dynamic focusing, unpolarized waves and beams, and many others). In addition to the steering angle (including zero degrees) of any physically generated wave or beam (such as the example above), soft steering of plane waves or dynamic focusing in transmission or reception (steering angle is zero degrees). This soft plane wave steering, among other things, can be thought of as augmenting physical transmit steering, purely steering any physically transmitted wave or beam, or receive dynamic focusing. In addition to the case where the angle φ is zero degrees, it can also be considered that reception steering is performed by software. The same applies to three-dimensional beamforming using a two-dimensional array. In addition, it is as having described in the method (1).

また、３次元の場合には、受信信号の３次元フーリエ変換Ｒ'(ｋ_ｘ,ｋ,ｋ_z)に対し、式（７'）と（８'）と共に、式（Ｃ４４）又は式（Ｄ４４）の波数マッチングを補間近似を通じて行い（式（Ｇ４４））、Ｆ(ｋ_ｘ',ｋ_y',ｋ_z')が３次元逆フーリエ変換される。この処理も、先行技術文献には開示されていない。

In addition, as in methods (1) and (2), physically and mathematically, wavenumber matching can be performed during the first Fourier transform, and wavenumber matching can be performed during the final inverse Fourier transform. is also possible. Depending on the steering, in transform processing involving wavenumber matching, the transform processing can be divided into directions and one-dimensional processing can be performed, thereby speeding up the processing. Also, among the wave number matching, the parts of the equations (F41) and (F43) and the equations (G41) and (G43) are instead frequency modulated by multiplying by the spatio-temporal signal using their complex exponential functions. (frequency modulation of kx, kz and ky, respectively).
Wavenumber matching in beamforming using equations (F41), (F42), and (F43) for two dimensions and equations (G41), (G42), and (G43) for three dimensions is performed through interpolation approximation. , you can get results quickly.
In the two-dimensional case, for the two-dimensional Fourier transform R′(k _x ,k) of the received signal, interpolate the wave number matching of Eq. (18b) or Eq. (18c) along with Eqs. (Equation (F44)), and F(k _x ', k _y ') is two-dimensionally inverse Fourier transformed. This approximation process is also not disclosed in the prior art document.

In the case of three dimensions, for the three-dimensional Fourier transform R'(k _x ,k,k _z ) of the received signal, the following equation (C44) or (D44) ) is performed through interpolation approximation (equation (G44)), and F(k _x ', ky ', k _z ') is _subjected to a three-dimensional inverse Fourier transform. This process is also not disclosed in the prior art documents.

Also, in a convex transducer, sector scan, IVUS, etc., when transmitting or receiving a wide wave in the direction of the angle θ in the polar coordinate (r, θ) in the direction of the radius r (cylindrical wave) ( FIG. 7 ), When transmitting a wide wave (cylindrical wave) in the direction of the angle θ in the same polar coordinate system (r, θ) as in FIG. 7 using a virtual source installed behind a physical aperture of arbitrary shape (see FIGS. 8A(a)-(c)) and echo signals collected for aperture plane synthesis in the polar coordinate system, Cartesian coordinates (x, y) are converted to polar coordinates ( r, .theta.), and an image signal can be generated in the Cartesian coordinate system (x, y) or polar coordinates (r, .theta.). In addition, a physical aperture element array represented by the polar coordinate system as described above or a physical aperture having an arbitrary aperture shape may be used to generate a plane wave for transmission or reception at an arbitrary distance position, and beam forming may be performed in the same way (Fig. 8B(d)-(f)). This is equivalent to generating a virtual linear aperture array (or plane wave) at that distance position, and if the distance position is set to zero, it corresponds to the use of a virtual linear aperture array. Distance positions can also be set in front of the physical aperture in addition to behind it, and virtual linear aperture arrays (or plane waves) can also be generated at those positions. A virtual linear aperture array may be used as a virtual receiver instead of a virtual source, or may also serve as a virtual source. The same is true when other transmit beamforming is performed and when they are performed in a spherical coordinate system. On the other hand, similarly, when using a convex transducer, sector scan, IVUS, etc., or when using a virtual source installed behind a physical aperture with an arbitrary aperture shape, directly according to method (5) An image signal can be generated in Cartesian coordinates. In addition, a physical aperture element array represented by the polar coordinate system as described above or a physical aperture having an arbitrary aperture shape may be used to generate a plane wave for transmission or reception at an arbitrary distance position, and beam forming may be performed in the same way (Fig. 8B(d)-(f)). This is equivalent to generating a virtual linear aperture array (or plane wave) at that distance position, and if the distance position is set to zero, it corresponds to the use of a virtual linear aperture array. Distance positions can also be set in front of the physical aperture in addition to behind it, and virtual linear aperture arrays (or plane waves) can also be generated at those positions. A virtual linear aperture array may be used as a virtual receiver instead of a virtual source, or may also serve as a virtual source. In these cases, echo signal imaging, displacement measurement, etc. can be performed consistently in the same Cartesian coordinate system. The same applies to the polar coordinate system. In these, similarly to the method (1), it can be similarly processed in an arbitrary orthogonal coordinate system or converted into an arbitrary orthogonal coordinate system. In these, transmission focusing may be performed. In addition, as described above, the virtual source and the virtual receiver are not necessarily installed behind the physical opening, but may be installed in front of the opening. (Patent Document 7, Non-Patent Document 8). Thus, the present invention is not limited to the above. Also, in the wavenumber matching in these beamformings, an interpolation approximation process is sometimes performed to obtain an approximate solution at high speed.

In addition, the apodization process for transmission or reception or both of the received signals can be performed at various timings (by hardware during reception or by software after reception) because of linear processing. As described above, physical apodization may occur during transmission. It should be noted that in the case of performing wave number matching through interpolation approximation using Equation (7) or Equation (7′), in order to increase the accuracy, processing is performed under appropriate oversampling at the expense of an increase in the amount of calculation. There is a need. In that case, it should be noted that the number of data for Fourier transform increases unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed. These processes are performed in the same manner as in method (1), and are performed in the same manner in other methods (3) to (7).

Method (3): Multistatic Aperture Synthesis FIG. 11 is a schematic diagram of multistatic aperture synthesis. Multi-static aperture plane synthesis is a method in which ultrasonic waves are emitted from one element of an array and echoes are received by a plurality of elements around that element. A low-resolution image signal is obtained for each irradiation, and a high-resolution image signal is generated by superimposing a plurality of obtained low-resolution image signals. The present invention may be used to generate this low resolution echo signal.

As mentioned above, the conventional method is usually to generate low-resolution echo signals from the echo signals received for each emission of each element and superimpose them. On the other hand, in the present invention, a signal group consisting of signals having the same relationship of transmission and reception positions is treated as one set, digital monostatic aperture plane synthesis is performed for each set, and a plurality of signals are obtained by the present method (3). Finish the process by superimposing those low-resolution image signals. In practice, the superposition, which is a linear process, can be performed in the frequency domain before the horizontal inverse Fourier transform, which is faster, and the lateral alignment for the superposition is also When superimposing, by multiplying the complex exponential function for horizontal shifting processing and rotating the phase in the horizontal direction, the image signal can be generated at high speed without interpolation approximation. For the inverse Fourier transform, fast inverse Fourier transform should be performed once. Further, in order to generate each low-resolution echo signal, when applying the horizontal inverse Fourier transform, at the same time, a complex exponential function for performing horizontal shifting processing is applied to rotate the phase in the horizontal direction, and the spatial It is also possible to superimpose in regions. In that case, a dedicated inverse fast Fourier transform may be implemented.

However, what is important is that when s=2 and the distance in the x direction between the reception position at y=0 and the transmission position at y=0 is .DELTA.x when applying a program for monostatic aperture plane synthesis processing, For the distance y (coordinate y) of the point of interest, it is to calculate and use the converted distance y' of the propagation path to receive the signal at a position separated by Δx. ) is to calculate the reduced distance represented by When s=1, let Δx be the distance in the x direction between the receiving position at y coordinate y=0 and the transmitting position at non-zero y coordinate y=Y. When the angle is 0 degrees, it is to calculate the converted distance represented by equation (20b) (the y-coordinates of the transmitting position and the receiving position may be reversed).

When the deflection angle is θ (including zero degrees), multi-static aperture synthesis is performed to generate a beam subjected to at least reception dynamic focusing (when s=2, transmission dynamic focusing can also be achieved). As a beamforming method, waves are transmitted from each of a plurality of transmission aperture elements in a transmission effective aperture element array, and waves arriving from a measurement object are formed by at least one of a plurality of reception aperture elements at different positions. Receive and generate a received signal (even if there is one, it can be processed by the apparatus of the present invention), the transmitting aperture element is such that the wave motion is generated by at least reflection or backscattering in the object to be measured. (s=2), or at least those produced by transmission, forward scattering, or refraction in the measurement object (s=1), regardless of the x-coordinate of the receive aperture element that produces the receive signal. , and a y coordinate of zero, which doubles as any receive aperture element in the receive effective aperture element array or is different from any receive aperture element. One by one, or one by one of a plurality of transmit aperture elements in the transmit effective aperture element array opposite to the receive effective aperture element array having a non-zero constant y-coordinate (s= When 1, the y-coordinates of the transmission position and the reception position may be reversed).

That is, when steering is performed, the steering process described in method (2) is performed on each of the data groups for monostatic aperture plane synthesis generated by combining the transmitting elements and receiving elements at the same positions. and treat in the same way. The same is true if the steering angles for transmission and reception are different. However, even when the deflection angle is non-zero degrees, when applying the program for monostatic aperture plane synthesis processing, s =2 uses the conversion distance expressed by the formula (20a), and when s=1, the conversion distance expressed by the formula (20b) is used. Therefore, in the same way as methods (1) and (2), the deflection angle may be set to zero degrees or non-zero degrees in a deflectable program for processing.
For transmission, processing can be performed when plane waves are transmitted according to method (1), and processing can also be performed when arbitrary transmission beamforming such as fixed focusing is performed.

As another method, the y-coordinates of the transmission and reception aperture elements are set to zero for each of a received signal group obtained by arranging received signals having the same distance Δx in the horizontal coordinate between the transmission position and the reception position. between the transmit aperture element, the point of interest, and the receive aperture element, expressed using the y-coordinate of the point of interest, including the deflection angle θ and the positions of the transmit and receive apertures, and the distance Δx, when (s=2) (20c)), or when the y-coordinate of the transmission aperture element is non-zero (s = 1, the y-coordinates of the transmission position and the reception position may be reversed) , the distance between the transmitting aperture element, the point of interest, and the receiving aperture element expressed using the argument θ, the y coordinate of the point of interest, the y=Y coordinate of the transmitting aperture element, and the distance Δx (equation (20d)) is used to correct each of the image signals obtained by setting the deflection angle in the above monostatic aperture plane synthesis with respect to the horizontal position in the frequency domain, and superimpose them to obtain an image signal without interpolation approximation processing. can be generated. Large steering angles can be generated at the expense of spatial resolution in the depth direction.

Further, when performing three-dimensional wave digital signal processing using a two-dimensional aperture element array, for example, the axial direction y determined by the orientation of the aperture of the flat receiving aperture element array and the lateral directions x and z can be similarly treated in a Cartesian Cartesian coordinate system using the coordinates of , where the zero or non-zero deflection angle between the generated beam direction and the axial direction is expressed in terms of elevation angle θ and azimuth angle φ in which waves are transmitted from each of the plurality of transmission aperture elements in the transmission effective aperture element array, and waves coming from the measurement object are received by at least one of the plurality of reception aperture elements at different positions. The transmit aperture element produces a signal, the wave generated by at least reflection or backscattering in the measurement object (s=2), or by at least transmission, forward scattering, or refraction in the measurement object. has any x- and z-coordinates independent of the x- and z-coordinates of the receive aperture element generating the received signal, and a y-coordinate of zero, such that s=1. each one of a plurality of transmit aperture elements that doubles as any receive aperture element in the receive effective aperture element array or is different from any receive aperture element, or has a non-zero constant y-coordinate; Each of a plurality of transmission aperture elements in the transmission effective aperture element array located opposite to the reception effective aperture element array (when s=1, the y-coordinates of the transmission position and the reception position are reversed). may be used). When the deflection angle is zero degrees (no steering) or non-zero degrees (with steering), similar to the two-dimensional wave digital signal processing using the above one-dimensional aperture element array, the transmission element and the reception element Each data group for monostatic aperture plane synthesis generated from a combination of the same positions may be subjected to the processing with or without steering described in method (2), and the same processing may be performed. The same is true if the steering angles for transmission and reception are different. However, what is important is that when s = 2, y If Δx is the distance in the x direction (horizontal direction) and Δz is the distance in the z direction (elevation direction) between the reception position of =0 and the transmission position of y = 0, the reception is at positions separated by Δx and Δz in two directions. is to calculate and use the reduced distance y' of the propagation path to , and to calculate the reduced distance represented by equation (20e). When s=1, the distance in the x direction between the receiving position at y coordinate y=0 and the transmitting position at non-zero y coordinate y=Y is Δx, and the distance in the z direction is Δz. Calculating the scaled distance represented (the y-coordinates of the transmitting and receiving positions may be reversed).

For transmission, processing can be performed when plane waves are transmitted according to method (1), and processing can also be performed when arbitrary transmission beamforming such as fixed focusing is performed.

As another method, for each of a received signal group obtained by arranging received signals having equal distances Δx and Δz in the horizontal x-coordinate and z-coordinate between the transmission position and the reception position, Transmit aperture expressed using the y-coordinate of the point of interest, including the elevation angle θ and azimuth angle φ, transmit and receive aperture positions, and the distances Δx and Δz when the y-coordinate of the aperture element is zero (s=2) Half the linear distance between the element, the point of interest, and the receiving aperture element, or the y-coordinate of the transmitting aperture element is non-zero (s=1, the y-coordinates of the transmitting and receiving positions are and azimuth angle φ, the y coordinates of the point of interest and the y coordinates of the transmit aperture element, and the distances Δx and Δz between the transmit aperture element, the point of interest, and the receive aperture element when each of the image signals obtained by setting the deflection angle in the above-mentioned monostatic aperture plane synthesis is corrected with respect to the horizontal position in the frequency domain using the distance of image signal can be generated without Large steering angles can be generated at the expense of spatial resolution in the depth direction.

Also, in order to generate an image signal representing the propagation of the unknown wave source or the wave generated by it (so-called passive mode), the y-coordinate of the estimated unknown wave source is set to the y-coordinate of the transmitting aperture element, Beamforming is good. It is also effective to observe while changing the y-coordinate by trial and error. For example, it is preferable to obtain effects such as imaging, higher spatial resolution, stronger signal intensity, and higher contrast. Using these as criteria, a series of processes can be automatically performed. is also possible.

As will be explained later, the information about the wave source position or transmit aperture element may be given the position relative to the receive aperture element, the direction or distance of the position to be located, the direction of the aperture, or the direction of propagation of the generated wave. Also, the time at which the wave was generated by any wave source may be given. It may be observed by other devices, or it may be transmitted from the source of the received signal itself or from a wave propagating at a higher speed.

For the received signal, find the centroid (central) frequency or instantaneous frequency of the multidimensional spectrum, find the direction in which the wave source exists or the propagation direction of the wave, adjust the deflection angle of transmission or reception, and obtain the beam Forming may also be performed. In addition, for the beamformed image signal, the centroid (central) frequency or instantaneous frequency of the multidimensional spectrum is obtained, and from this, the direction in which the wave source exists or the wave propagation direction is obtained, and the deflection angle of transmission or reception is adjusted to perform beamforming. These processes can be performed at a plurality of reception apertures or effective reception apertures to geometrically determine the position or direction of the wave source. These processes are useful and may have other beamforming applications.

As explained in method (2) for monostatic aperture plane synthesis, in multistatic aperture plane synthesis for method (3) as well, arbitrary beamforming can be generated using echo signals collected for aperture plane synthesis. (Actually, image signals can also be generated by performing the processing described in methods (1) and (4) to (7) on those data). It may be effective to have a large amount of data compared to the monostatic type, but the amount of calculation increases. In the plane wave processing of the method (1) as well, the aperture plane synthesis processing can be performed by using the code. Also, in a convex transducer, sector scan, IVUS, etc., when transmitting or receiving a wide wave in the direction of the angle θ in the polar coordinate (r, θ) in the direction of the radius r (cylindrical wave) ( FIG. 7 ), When transmitting a wide wave (cylindrical wave) in the direction of the angle θ in the same polar coordinate system (r, θ) as in FIG. 7 using a virtual source installed behind a physical aperture of arbitrary shape (see FIGS. 8A(a)-(c)) and echo signals collected for aperture plane synthesis in the polar coordinate system, Cartesian coordinates (x, y) are converted to polar coordinates ( r, .theta.), and an image signal can be generated in the Cartesian coordinate system (x, y) or polar coordinates (r, .theta.). The same is true when other transmit beamforming is performed and when they are performed in a spherical coordinate system. On the other hand, it is also possible to generate image signals in Cartesian coordinates directly according to method (5), such as in convex transducers, sector scans, or IVUS. In that case, echo signal imaging, displacement measurement, etc. can be performed consistently in the same Cartesian coordinate system. In these, similarly to the method (1), it can be similarly processed in an arbitrary orthogonal coordinate system or converted into an arbitrary orthogonal coordinate system. In addition, the beamforming described in paragraphs 0211 to 0222 of method (1) and paragraph 0240 of method (2) can be performed in the same manner. It can be installed arbitrarily regardless of the shape (Patent Document 7, Non-Patent Document 8). Moreover, the present invention is not limited to this (the same shall apply hereinafter).

Also, as noted above, it is possible to alternately deflect to produce large steering angles at the expense of spatial resolution in the depth direction. different steering angles can be handled in the same way. The converted distance may be calculated using the lateral distance between the transmission aperture element and the reception aperture element, the transmission steering angle and the reception steering angle, and the distance between the transmission aperture element and the reception element in the case of a transmissive type.

Any steering in method (3) is essentially a software implementation. Also, apodization may or may not be performed at the time of transmission. In addition, since the reception apodization process is also a linear process, it can be performed at various timings (in terms of hardware or software). This can be easily performed at an appropriate timing in consideration of the amount of calculation that depends on the effective aperture width and the like that determine the number of . For example, each set can be weighted to initiate the generation of low-resolution echo signals, or the low-resolution signals generated can be apodized in the frequency or spatial domain.

In addition, in applying the monostatic aperture plane synthesis of method (2), it is possible physically and mathematically to perform wavenumber matching during the first Fourier transform, and to perform wavenumber matching during the final inverse Fourier transform. It is also possible to Also, wavenumber matching may be performed at high speed through the interpolation approximation described above. Interpolation approximation may be linear interpolation approximation or approximation using the nearest data itself, high-order interpolation approximation may be performed, or a sinc function may be used. To improve the accuracy of wavenumber matching through interpolation approximation, it is necessary to perform processing under appropriate oversampling at the expense of an increase in the amount of calculation. In that case, it should be noted that the number of data for Fourier transform increases, unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed.
In addition, in the horizontal alignment for superimposition described in paragraphs 0243, 0247, and 0250, the phase is rotated in the horizontal direction by multiplying by a complex exponential function for performing horizontal shifting processing. Alternatively, spatial shifting operations may be performed through interpolation approximations to achieve faster processing. Interpolation approximation may be linear interpolation approximation or approximation using the nearest data itself, high-order interpolation approximation may be performed, or a sinc function may be used. Even in this case, in order to improve the accuracy of the interpolation approximation, it is necessary to perform processing under appropriate oversampling at the expense of an increase in the amount of calculation.

Method (4): Fixed Focusing FIG. 12 is a schematic diagram of fixed focusing using a linear array. Fixed focusing is a method in which one point is set as a focus point and a delay is given to each element so that ultrasonic waves reach the focus point at the same time. A measurement target is scanned by receiving part or all of the physical aperture of the array type transducer as an effective aperture. Of course, steering may also be performed. The transmit and receive steering angles may be different.

Fixed focusing is the generation of an image signal, method (1): beamforming during plane wave transmission, or method (3): multi-static aperture synthesis, or beamforming and method for plane wave transmission of method (1). (2) or a method combining the receiving dynamic focusing of method (3). In that case, there are the following three methods.
(i) One image signal generating process is performed on each superimposed received signal obtained in the effective aperture width.
(ii) Generate a so-called normal low-resolution image signal using the received signal for each transmission and superimpose them.
(iii) Similar to multi-static aperture plane synthesis, a set of signals having the same positional relationship for transmission and reception is used to generate image signals, which are superimposed.

Cartesian coordinates ( x, y) can be converted to polar coordinates (r, θ) for processing, and an image signal can be generated at the polar coordinates (r, θ). As noted above, an interpolation approximation may be required after the image is generated. The same is true for transmission and reception using a spherical coordinate system. There is a report (Non-Patent Document 6) of processing that includes approximation processing in the case of transmission focusing, and similarly results are obtained in polar coordinates (r, θ). The inventors of the present application have proposed a method (5) for generating image signals directly in the Cartesian coordinate system as a result of transmit or receive beamforming in their polar or spherical coordinate system, or any Cartesian curvilinear coordinate system. , (5-1), (5-1′), and (5-2) were also invented. In that case, imaging of transmitted signals, reflected signals, scattered signals, attenuation signals, displacement measurements, etc., can be performed consistently in the same Cartesian coordinate system. In these, similarly to the method (1), it can be similarly processed in an arbitrary orthogonal coordinate system or converted into an arbitrary orthogonal coordinate system. In addition, the beamforming described in paragraphs 0211 to 0222 of method (1) and paragraph 0240 of method (2) can be performed in the same manner. It can be installed arbitrarily regardless of the shape (Patent Document 7, Non-Patent Document 8). Moreover, the present invention is not limited to this (the same shall apply hereinafter). In addition, although deflection is possible as described above, it is also possible to handle the case where the deflection angles of physical transmission beamforming and soft reception beamforming are different. Transmit steering can also be applied in software. In that case, the deflection angle may differ from other deflection angles. It is also possible to physically beamform when receiving. Sending and receiving can also be interpreted inversely. Apodization may be performed during transmission, and reception apodization processing may be performed on the received signal (by hardware during reception or by software after reception). When apodization is performed by software, it is performed according to method (1) or method (3). Apodization is also performed in a similar manner when using a method combining beamforming for plane wave transmission and reception dynamic focusing of method (2) or method (3).

It should be noted that any process of method (4) using method (1) with plane wave processing is theoretically and practically capable of beamforming any physical transmit or receive; By processing as described above, various combinations of beamforming can be performed (for example, physical beamforming using a computer, a dedicated device, etc., which is different from soft beamforming performed using a computer, a dedicated device, etc.) Each or both of the transmit and receive processes, including focusing, steering, apodization, etc., which may occur at those times, can be processed as plane wave transmit and receive, or as described above, transmit and receive can be reversed. can be captured and processed). For simultaneous transmission of multiple beams, e.g., as described in paragraphs 0109, 0112, 0365, 0367, 0368, etc., the multiple beams may or may not interfere with each other, with or without physical deflection. physical focus (sub-aperture width, distance or depth, position, etc.) are the same or different, and regardless of whether the physical transmission deflection angles are the same or different, the above processing of method (4) is valid, in particular, (i ), a high frame rate can be achieved by using a method in which one image signal generating process is applied to superposition of received signals obtained at the effective aperture width of . In method (4), it is not necessary to perform beamforming at positions where waves do not interfere (described in paragraphs 0030, 0364, etc.). A high frame rate can be achieved in processing. In this case, if the soft transmission deflection angles and the reception deflection angles applied to the plurality of focus beams to be executed are the same, the above processing may be performed. In addition, even in the case of performing multiple transmissions such as focusing at multiple positions or transmission dynamic focusing at the same phase of the target, the received signals can be superimposed in the same manner. This processing can be applied to all of the reception signals received in the same phase of interest, if the reception signals are superimposed in a state where the times are aligned based on the position and timing of the transmitting element.

If any of the soft transmission deflection angles and the reception deflection angles applied to a plurality of focus beams is different, they are divided into identical ones, the same processing is applied to each identical one, and the final result is obtained. It suffices to perform superposition in the frequency domain in order to obtain it. For one physical transmit beam (with or without deflection), multiple polarized receive beams (including 0° deflection angle) may be generated and similarly processed. Even when a plurality of different physical deflections are performed, they may be divided into the same ones to obtain respective processing results, or they may be processed without division. If separated, the superposition may be done in the spatial domain or the frequency domain.

When physically transmitting in multiple directions, a unique transmission and reception deflection may be applied to each transmission deflection angle in terms of software. Alternatively, separation processing such as independent component analysis (ICA: there are many references, including Te-Won Lee, Independent Component Analysis: Theory and Applications, Springer, 1998, which is a relatively old book). may be applied and processed. Analog devices may also be used. An example is to set the soft transmit and receive deflection angles to be the same as their respective physical deflection angles. Signal separation is described elsewhere (eg, paragraph 0370).

Although the method is described herein as being implemented for various fixed focusing operations, the method is not limited to these and can be used when other transmit beamforming is implemented. With or without focusing (multi-focus that realizes multiple different focus positions with respect to the effective aperture), with or without deflection (multiple steering with different deflection angles), with apodization (including different cases for each position) Or none, even when performing multiple transmissions or receptions with different waves and ultrasonic parameters such as different F numbers, different transmission ultrasonic frequencies or transmission bands, different reception frequencies or reception bands, different pulse shapes, different beam shapes, etc. , can be similarly processed, and beamforming can be performed with new features that cannot be generated by a single beamforming with them fixed. For example, it is well known that superimposition can obtain multiple focal points and broaden the bandwidth (higher resolution) both in the depth direction and in the lateral direction, and these processes can be performed at high speed. If, for example, the so-called pulse inversion method (radiating pulses with different polarities as ultrasound parameters) is used to obtain harmonics, the received signals can be superimposed and likewise processed at high speed. Of course, the received signals can also be superimposed after performing beamforming. The received signals of two or more beams may be superimposed.

On the basis of the transmission and reception reversed above, the above process may also be applied to simultaneous receive beamforming. Also, the above processing may be applied to both transmission and reception.

In addition, when processing is divided into a plurality of beamformings, parallel processing may be performed. It may be divided into a plurality of beam formings according to various waves such as the above deflection angle, ultrasonic parameters, the position of the region of interest, and the like. A single received signal may be used for multiple purposes, such as imaging, measurement, and therapy. Signals may be generated by forming and post-processing such as filtering may be used to generate a signal suitable for the purpose, but appropriate beamforming may be performed according to the purpose and they may be processed in parallel.

Beamforming when transmitting arbitrary waves (including non-beamformed waves) in addition to arbitrary beams such as fixed focus beams, superposition processing when transmitting multiple beams and waves, and simultaneous multiple beams and processing when transmitting waves. That is, receive beamforming (dynamic focusing, etc.) can be performed at once for any single or multiple transmissions. Multiple beamforming may be performed using multi-directional aperture synthesis, which is also fast. Moreover, the present invention is not limited to them.
Physically and mathematically, wave number matching can be performed during the first Fourier transform, and wave number matching can be performed during the final inverse Fourier transform. One of the features of the present invention is that no interpolation approximation is performed in wave number matching. , Interpolation approximation processing may be performed in wavenumber matching as in methods (1) to (3), and approximate wavenumber matching described therein is performed to perform high-speed beamforming. There is In order to perform approximate wavenumber matching with high accuracy, it is necessary to appropriately oversample the received signal at the expense of an increase in the amount of calculation. In that case, it should be noted that the number of data for Fourier transform increases, unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed.

Method (5): Image signal generation in polar coordinate system Method (5) transmits or receives (a part of) an ultrasonic cylindrical wave in a two-dimensional polar coordinate system (r, θ) such as a convex array, sector scan, or IVUS. Fig. 7 is a method of generating an image signal in the Cartesian coordinate system for the case (see Fig. 7). Methods (1)-(4) and (6) can be implemented.

極座標系における受信エコー信号は、f(r,θ)と表されるので、式（２３）が成立する。

従って、ヤコビ（Jacobi）演算を通じて、式（２４）が得られる。この様にして、極座標系において表される波動をデカルト座標系において平面波成分(k_x,k_y)に分解できる。任意の直交曲線座標において表される波動も同様に、デカルト座標系において平面波成分(k_x,k_y)に分解できる。

Polar coordinate representation of the Fourier transform will be described below. A two-dimensional Fourier transform is represented by Equation (22).

Since the received echo signal in the polar coordinate system is expressed as f(r, θ), Equation (23) holds.

Therefore, equation (24) is obtained through Jacobi arithmetic. In this way, a wave represented in a polar coordinate system can be decomposed into plane wave components (k _x , k _y ) in a Cartesian coordinate system. A wave represented in arbitrary Cartesian curvilinear coordinates can likewise be decomposed into plane wave components (k _x , k _y ) in the Cartesian coordinate system.

ここで、ｒ_０はコンベックス探触子の曲率半径であり、x₀とy₀はコンベックス探触子のアレイ素子位置を表すx軸とy軸の座標である。ステップＳ２１において、受信信号を時間ｔに関してフーリエ変換し、ステップＳ２２において、受信信号を角度θに関してフーリエ変換（ＦＦＴ）することにより、極座標系において受信した信号をデカルト座標系において平面波成分(k_x,k_y)に分解できる。 Method (5-1): Image Signal Generation for Cylindrical Wave Transmission or Reception FIG. 13 is a flow chart showing digital signal processing during cylindrical wave transmission. From Equation (24), Fourier transform in the direction of angle θ on the aperture is expressed by Equation (25).

Here, r ₀ is the radius of curvature of the convex probe, and x ₀ and y ₀ are the x-axis and y-axis coordinates representing the array element positions of the convex probe. In step S21, the received signal is Fourier transformed with respect to time t, and in step S22, the received signal is Fourier transformed (FFT) with respect to angle θ, thereby transforming the signal received in the polar coordinate system into the plane wave components (k _x , ) in the Cartesian coordinate system. k _y ).

Therefore, for example, an image signal can be generated by subjecting this to wave number matching represented by equation (26) (step S23) and inverse Fourier transform with respect to the space (x, y).

若しくは、ステップ２３とステップ２４とを逆にして演算しても良い。 Further, in step S24, the two-dimensional spectrum is multiplied by the following complex exponential function to calculate the angular spectrum at each depth y.

Alternatively, the calculation may be performed by reversing steps 23 and 24 .

Alternatively, without using equations (26a) and (26b), according to method (5), multiply by the following complex exponential function, perform wavenumber matching, and obtain the angular spectrum at each depth position y good.

Furthermore, for example, the image signal is obtained in step S27 by summing the frequency components k of the angular spectrum in step S25 and performing an inverse Fourier transform (IFFT) on the horizontal wavenumber kx in step S26. A pure two-dimensional inverse Fourier transform may be applied.

When steering is performed, wave number matching in the x and y directions and spatial resolution can be obtained in accordance with method (1) and equations (9a) to (9c) using the deflection angle θ. As will be described later, when calculation is performed in the polar coordinate system (r, θ), a steering angle (angle formed with the radial direction) can be set in the polar coordinate system and steering can be performed in the same manner. Physical steering can be implemented, transmission or reception, or soft steering of transmission and reception can be implemented, or a combination of physical steering and soft steering can be implemented. There are like other methods, such as method (1).

This method does not require interpolation processing for wave number matching and coordinate system conversion in obtaining image signals in the Cartesian coordinate system (x, y) from signals in the polar coordinate system (r, θ), and is fast and highly accurate. It performs beamforming. Similar to plane wave transmission in linear array transducers, cylindrical waves can also be deflected in a polar coordinate system. The same processing can be applied to the case where the steering angles of transmission beamforming and reception beamforming are different. Soft steering can also be applied. Apodization can be performed as well. In the case of cylindrical waves, at different positions in the z-axis direction orthogonal to the two-dimensional polar coordinate system (that is, the z-axis in the cylindrical coordinate system (r, θ, z)), the above transmission is performed simultaneously and received, In some cases, the above transmissions are performed in the same time phase but at different times, and the received signals are superimposed and the above processing is performed. The z-axis direction may be focused by an analog device (lens), and arbitrary processing may be performed by the digital signal processing of the present invention. It can be similarly calculated when the wave propagation direction is in the central direction (useful for, for example, HIFU treatment, various imaging with a circular-based array transducer surrounding the object, CT, etc.). Of course, receive-only beamforming may also occur in them and is handled similarly. It should be noted that the received signal represented by polar coordinates (r, θ) can be processed by replacing the Cartesian coordinates (x, y) in method (1) with polar coordinates (r, θ). , .theta.) can be generated as described above, and interpolation approximation will be performed in subsequent processing. In these, steering can be performed as well. Methods (2) to (4) and (6) can also be carried out in the same manner.

Also, when the received signal is expressed as a digital signal of Cartesian coordinates (x, y), f(x, y) is Fourier transformed with respect to the radius r and the angle θ to be processed, contrary to Equation (22). , it turns out that the image signal is generated in the polar coordinate system (r, θ), or each method can also be used to generate the image signal in the Cartesian coordinate system (x, y). Steering and apodization may be performed as well.

Further, as shown in FIGS. 8B(d) to (f), using a physical aperture element array represented by the polar coordinate system as described above or a physical aperture having an arbitrary aperture shape, transmission or reception can be performed at an arbitrary distance position. Alternatively, both plane waves may be generated and beamforming may be performed in the same way, and the image signal is generated in a Cartesian or polar coordinate system, or in a Cartesian polar coordinate system set to match the physical aperture shape. There is something. This is equivalent to generating a virtual linear aperture array (or plane wave) at that distance position, and if the distance position is set to zero, it corresponds to the use of a virtual linear aperture array. Distance positions can also be set in front of the physical aperture in addition to behind it, and virtual linear aperture arrays (or plane waves) can also be generated at those positions. The plane wave may be steered or the virtual linear aperture may be tilted (virtually mechanically steered). Of course, the physical aperture is mechanically steered if desired. Their transmit-only or receive-only cases are based on the case of transmit and receive cylindrical waves, respectively, and sometimes other beamforming may also be performed.

Method (5-1'): Image signal generation using virtual source and other arbitrary shape aperture array The case of generating a portion of a cylindrical wave using an installed virtual source (see FIGS. 8A(a)-(c)) will be described.

(i) When using the received signal acquired for monostatic aperture plane synthesis, if necessary, the complex exponential function is applied to the Fourier transform of the received signal transmitted and received by each element and stored in a memory or the like. Multiply and take the wave response emitted from the virtual source as a digital received signal represented by polar coordinates (r, θ), and then follow method (5) or use method (5-1) , can directly generate an image signal in the (x,y) coordinate system. Similarly, it is also possible to express the received signal as a digital signal of polar coordinates (r, θ), and perform processing by replacing the Cartesian coordinates (x, y) in method (1) with polar coordinates (r, θ). It is also possible to generate an image signal at (r, θ). Of course, the monostatic processing of method (2) is also possible in the polar coordinate system (r, θ).

(ii) When using received signals acquired for multi-static aperture plane synthesis, the received signals transmitted by each element, received by surrounding elements, and stored in a memory or the like may be processed, if necessary. For example, the Fourier transform is multiplied by a complex exponential function, the wave response emitted from the virtual source is taken as a digital received signal represented by polar coordinates (r, θ), and the multi-static aperture plane synthesis of method (3) is performed. can. As another process, according to method (5), or superimpose the digital received signals in each receiving element, and then use method (5-1) directly (x, y) coordinate system can generate an image signal at Similarly, the digital received signals can be superimposed in each receiving element, and then processed by replacing the Cartesian coordinates (x, y) in method (1) with polar coordinates (r, θ). It is also possible to generate an image signal at (r, θ). Of course, without superposition, the multistatic processing of method (3) can also be performed in the polar coordinate system (r, θ).

(iii) In order to eliminate the process of rewriting the signals received by the physical aperture array into digital signals in the polar coordinate system (r, θ) in these processes, the sampling is originally performed in the polar coordinate system (r, θ). A transmit or receive delay pattern may be used to transmit and receive sample from each aperture element. Then, the image signal can be directly generated in the (x, y) coordinate system based on the method (5-1) or the method (2) or (3) based on the method (5). Similarly, the received signal is represented as a digital signal of polar coordinates (r, θ), and the Cartesian coordinates (x, y) in methods (1) to (3) are converted to polar coordinates (r, θ). can also be used to generate an image signal in polar coordinates (r, θ).

(iv) Similarly, in the case of generating a part of a cylindrical wave using a virtual source installed behind an arbitrary aperture shape (see FIGS. 8A (a) to (c)), the above (i) to ( In iii), the Fourier transform of the received signal received by each element and stored in a memory or the like is multiplied by a complex exponential function to express the received signal as a digital signal of Cartesian coordinates (x, y) (however, time necessary) or interpolation approximation, and inversely to formula (22) in method (5-1), f(x, y) is Fourier-transformed with respect to radius r and angle θ, and finally, polar coordinates Either the image signal is generated in the system (r, θ), or each method can be used to generate the image signal in the Cartesian coordinate system (x, y). An image signal can also be generated in a similar manner in an orthogonal curvilinear coordinate system (curvilinear coordinate system) set in accordance with the aperture shape.

In addition, in (i) to (iv), various beamforming and the like described in method (5-1) can be performed.

Incidentally, as described in method (1), etc., a virtual source (Fig. 8A ( a) to (c)) is used to transmit a cylindrical wave (using a transmission delay), each element is coded and transmitted in the same manner as in the case of plane wave transmission, and then received Decoding the received signal to generate a received signal group for aperture plane synthesis, and directly generating an image signal in an arbitrary orthogonal curvilinear coordinate system such as a Cartesian coordinate system or a polar coordinate system by the above processing. can be done. Also, a virtual receiver may be set instead of the virtual source, and the virtual receiver may also serve as the virtual source.

In addition, a virtual source (Fig. 8A ( a) to (c)) is used to transmit a cylindrical wave (using a transmission delay), the above method can be used to:
(A) Method (1) is directly applied to the received signal represented by the Cartesian coordinate system (x, y) obtained by the linear array type transducer or mechanical scanning, and the image is imaged in the Cartesian coordinate system. get the signal.
(B) Multiplying the frequency response obtained by the (fast) Fourier transform in the y direction for the signal received at each receiving position by a linear array type transducer or mechanical scanning, and multiplying the complex exponential function in the y direction Shifting, in the polar coordinate system (r, θ) with the virtual source as the origin, corrected to the position coordinate in the radial r direction under θ determined by the reception position, and applying method (5) or method (5-1) , Cartesian coordinates (x, y) or polar coordinates (r, θ). Instead of spatial shifting using a complex exponential function, approximate spatial shifting by zero padding of signal values in the r coordinate system is sometimes performed. , requiring a high sampling rate AD converter and a large amount of memory, and it should be noted that the number of data increases before the Fourier transform.
(C) Applying method (5) or method (5-1) to signals received in an arbitrary aperture shape other than a linear array transducer or a pseudo linear array aperture by mechanical scanning, and similarly Cartesian An image signal can be similarly generated in a coordinate system (x, y), a polar coordinate system (r, θ), or an orthogonal curvilinear coordinate system (curvature coordinate system) set in accordance with the shape of the aperture.
(D) A virtual receiver may be used instead of a virtual source, and a virtual receiver may double as a virtual source.

As a result of these methods (5-1'), for example, using another type of transducer or another mechanical scan, a convex or sector type transducer as shown in FIG. ) can be generated in the Cartesian coordinate system (x, y), the polar coordinate system (r, θ), or the orthogonal curvilinear coordinate system (curvilinear coordinate system) set according to the shape of the aperture. .
In addition, conversely, when a linear type transducer is virtually used by using a physically different type transducer (for example, when a physically convex type transducer is used, FIG. 8B (d) to (f) ), if the virtual source or virtual receiver is located at or behind or in front of the physical aperture), similarly in the Cartesian coordinate system (x, y) or in the polar coordinate system (r, θ) can also be generated in an orthogonal curvilinear coordinate system (curvilinear coordinate system) set in accordance with the shape of the opening.
As a special case, for example, in the case of physically using a linear array type transducer, a case of generating a cylindrical wave using a virtual source or a virtual receiver behind the physical aperture is applied, and lateral It is also possible to generate an image signal in the case of generating a plane wave or a virtual linear array type transducer (FIG. 8B(g)) spread out in a direction.
In these, the generated transmit or receive waves may be steered or virtually aperture tilted (virtually mechanically steered). Of course, the physical aperture is mechanically steered if desired.

Method (5-2): Image Signal Generation at Fixed Focus FIG. 14 is a schematic diagram of fixed focusing using a convex array. Fixed focusing can also be achieved in convex arrays as shown in FIG. FIGS. 14A and 14B are schematic diagrams of fixed focusing positions set at equal distances from each effective aperture and at arbitrary distances from the convex array, respectively. As in the case of the linear array type (method (4)), the image signal can be generated by the same calculation processing as in the case of cylindrical wave transmission. That is, based on the processing of method (1) or method (3), there are the following three methods.
(i) Each received signal obtained in the effective aperture width is subjected to image signal generation processing once with respect to superimposition.
(ii) Generate a so-called normal low-resolution image signal using the received signal for each transmission and superimpose them.
(iii) Similar to multi-static aperture plane synthesis, a set of signals having the same positional relationship for transmission and reception is used to generate image signals, which are superimposed.
Although the image signal can be generated directly in the Cartesian coordinate system as described above, it is also possible to implement method (4) using the coordinate axes of the polar coordinate system to generate the image signal in the polar coordinate system. Deflection and apodization can be performed as well. The z-axis direction can be processed in the same manner as in (5-1).

Also, when the received signal is represented as a digital signal with Cartesian coordinates (x, y), f(x, y) is processed by Fourier transform with respect to radius r and angle θ, ultimately resulting in a polar coordinate system (r , θ), or each method can also be used to generate image signals in the Cartesian coordinate system (x,y). Steering and apodization, z-axis processing may be performed as well.

When steering is performed, wave number matching in the x and y directions and spatial resolution may be obtained according to the method (4). As will be described later, when calculation is performed in the polar coordinate system (r, θ), a steering angle (angle formed with the radial direction) can be provided in the polar coordinate system and steering can be performed in the same manner. Physical steering can be implemented, transmission or reception, or soft steering of transmission and reception can be implemented, or a combination of physical steering and soft steering can be implemented. There are like other methods, such as method (1).

In addition, when using a virtual source or a virtual receiver, a virtual aperture is realized using the physical aperture described in method (5-1') or the like at that position or before and after the above. Transmit fixed focusing can be done. For example, a linear array transducer may be virtually implemented. Others may implement transducers with arbitrary aperture shapes. Image signals are generated in a Cartesian or polar coordinate system, or in a Cartesian Cartesian coordinate system set to match the physical aperture geometry. Physical steering can be implemented, transmission or reception, or soft steering of transmission and reception can be implemented, or a combination of physical steering and soft steering can be implemented. There are like other methods, such as method (1). In these, the generated transmitted or received waves may be steered or virtually aperture tilted (virtually mechanically steered). Of course, the physical aperture is mechanically steered if desired.

As described above, the beamforming of methods (1) to (4) can be performed, but the same effect can be obtained by adapting to arbitrary beamforming without being limited to them. In particular, when method (4) is used, receive beamforming can be performed for any transmit beam or wave in addition to the transmit fixed focus beam. Of course, beamforming of the collective received signal and the superimposition of the received signal for each transmission during simultaneous transmission of multiple different beams or waves can be performed as well.

Method (5-3): Image signal generation during reception in a spherical coordinate system When using a spherical core-shaped wave aperture element array, three-dimensional digital wave signal processing is performed. If the element array is so, the reception of the wave takes place in the spherical coordinate system (r, θ, φ), so the received signal of the received wave is denoted f(r, θ, φ). In this case as well, various beamforming can be performed through Jacobi operations, as in the case of the two-dimensional polar coordinate system (r, θ).

Specifically, a three-dimensional Fourier transform is performed on the received signal f(r, θ, φ) to decompose the received wave into plane waves in the Cartesian coordinate system (x, y, z), and the Cartesian coordinate system ( (27) expressed in the wavenumber domain or the frequency domain (kx, ky, kz) of _x , _y , _z ) is replaced by Jacobi By calculation, it is possible to calculate as in Equation (28) and directly generate an image signal in the Cartesian coordinate system without performing an interpolation approximation process. Of course, the beamforming of methods (1) to (4) and (6) can be performed, but the method is not limited to them, and any beamforming can be adapted and used to obtain the same effect. In particular, when method (4) is used, reception beamforming can be performed for all transmission beams or waves in addition to the transmission fixed focus beam, as in the two-dimensional case. Of course, beamforming of collectively received signals and superimposition of received signals for each transmission during simultaneous transmission of multiple different beams or waves can be performed as well. In addition, the use of virtual sources and virtual receivers, steering, etc., can all be implemented in the same way as in the two-dimensional case, using a Cartesian coordinate system or a spherical coordinate system, or in accordance with the physical aperture shape. An image signal can be generated in a set orthogonal polar coordinate system.

Method (5"): Image signal generation in arbitrary orthogonal curvilinear coordinate system when transmitting or receiving in Cartesian coordinate system. An image signal represented by a two-dimensional polar coordinate system or a spherical coordinate system can be obtained directly from the signal without interpolation approximation, and can be realized by similar calculations.For example, if the received signal is f(x, y, z), Fourier transform is performed in the r, θ, and φ directions, and the received signal is decomposed into circular waves and spherical waves, which correspond to plane waves in the Cartesian coordinate system, through Jacobian arithmetic. Good, these methods may also be used to vary the FOV (e.g., it may be widened) Jacobian arithmetic can be used to similarly generate image signals in any Cartesian coordinate system. , the image signal can be generated in any Cartesian coordinate system as well, even when transmitted or received in any coordinate system (different Cartesian coordinate systems, such as the Cartesian coordinate system and various curvilinear Cartesian coordinate systems). (including the case where the origin is different or rotated, and even if it is the same orthogonal coordinate system, the origin position is different or rotated).Similarly to the other methods described in method (5), Any transmit beam or wave can be processed, steering can be performed as well, and even virtual sources and virtual receivers can be used.

Physically and mathematically, wave number matching can be performed during the first Fourier transform, and wave number matching can be performed during the final inverse Fourier transform. One of the features of method (5) is that beamforming is performed in an arbitrary coordinate system without performing interpolation approximation in wavenumber matching. ~ When the beamforming described in method (4), method (6), and method (7) is performed in an arbitrary coordinate system, interpolation approximation processing may be performed in wave number matching, and each of them is described Approximate wavenumber matching may be performed to perform beamforming at high speed. In order to perform approximate wavenumber matching with high accuracy, it is necessary to appropriately oversample the received signal at the expense of an increase in the amount of calculation. In that case, it should be noted that the number of data for Fourier transform increases unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed.

Method (6): Migration Method In the apparatus of the present invention, migration processing can also be performed without performing interpolation approximation processing in wave number matching. The migration formula (formula (M6') below) itself is well known, and the derivation of the formula is also well known, so the derivation of the formula is omitted here.

Non-Patent Document 12 describes normal migration processing based on one-element reception by one-element transmission (that is, corresponding to undeflected processing using transmission/reception data for monostatic aperture plane synthesis in method (2)). As a basis, in the case of plane wave transmission and/or reception without and with steering (processing corresponding to method (1)), waves are generated from any transmit aperture element targeting any same position. Since the propagation time, which is the time until the wave is received by the reception aperture element that also serves as the transmission aperture element, is different from the case of normal migration, the propagation velocity and the coordinates of the target position are replaced (the following formula (M1 )), a method is disclosed to calculate the value expressed in the same form (Equation (M6)).

However, the processes of other methods (2) to (5) are not disclosed in Non-Patent Document 12 (not disclosed when steering is performed in method (2)). Furthermore, in calculating the formula (M6'), conventionally, interpolation approximation has been performed when performing wave number matching (formula (M4) and formula (M4')), but in the apparatus of the present invention, Wavenumber matching is performed with high accuracy without interpolation approximation (equation (M7) and equation (M7')).

Two-dimensional coordinates are taken with the horizontal direction as the x-axis and the depth direction as the y-axis, and the time coordinate is t. Specifically, in its normal migration, the propagation time required for a wave to make a round trip between arbitrary aperture element position (x,0) and arbitrary position (x _s , y _s ) is given by equation (M0) be done.

On the other hand, in a plane wave transmission with a steering angle θ (including 0°), the propagation time is given by equation (M0′).

Therefore, in the calculation based on the migration method when the polarized plane wave is transmitted in method (1), each of the carrier velocity c and the coordinate system (x _s , y _s ) representing the position of the object is replaced with the formula (M1). , the usual migration equations are calculated (equations (M4) and (M5)).

In summary, among the methods (1) to (5), migration is performed in the same manner except for the normal migration that performs undeflected processing using the transmission/reception data for monostatic aperture synthesis in method (2). can be processed. For example, the calculation procedure of the migration when transmitting a polarized plane wave (including 0°) in method (1) will be mainly described.

FIG. 15 is a flowchart showing migration processing when a polarized plane wave is transmitted. When the received signal is denoted as r(x,y,t), the received signal at the aperture element array position is denoted as r(x,y=0,t).

ここで、ｋ＝ω／ｃであり、波数ｋと角周波数（角振動数）ωは比例定数１／ｃで関係付けられ、１対１対応であり、ｋの代わりにωを用いて表したり計算できる。 First, as shown in equation (M2), the received signal is subjected to two-dimensional Fourier transform with respect to time t and horizontal direction x (two-dimensional fast Fourier transform is preferable).

Here, k=ω/c, the wave number k and the angular frequency (angular frequency) ω are related by a proportionality constant 1/c, and have a one-to-one correspondence, and are expressed by using ω instead of k. can be calculated.

As described above, a special two-dimensional fast Fourier transform method can also be used, but as a popular method, first, in step S31, for the received signal, at the horizontal coordinate x, at time t A Fast Fourier Transform (FFT) is performed on to obtain the spectrum of the analytic signal. Then, at each frequency coordinate within band k, a fast Fourier transform in the horizontal direction x may be performed (this is faster than calculating each two-dimensional spectrum according to formula (M2)).

If the plane wave is not transmitted with deflection, the above calculation is sufficient, but if it is deflected, trimming must be performed. Multiply the transformed R'(x,0,k) by the complex exponential function (M3). The operations can be computed directly at once, or dedicated Fast Fourier Transforms capable of such computations are also useful).

Then, in step S33, the received signal is subjected to fast Fourier transform (FFT) in the horizontal direction x. The result is represented here as R''(k _x ,0,k). By the way, even if the program is programmed to perform trimming, it is possible to process the case where a plane wave is transmitted without being deflected (steering angle 0°).

Wavenumber matching (or mapping) is then typically performed. In the case of beamforming methods (1) to (5) other than normal migration (method (2) without steering), propagation velocity c and coordinate system (x _s , y _s ) are expressed by the above equation Like (M1), the propagation velocity (E1) and the coordinate system (E2) for each beam forming are read and processed.

但し、受信信号が反射信号の場合には、s = 2であり、透過信号の場合には、s = 1である。
但し、式（Ｍ４）と式（Ｍ４'）の波数マッチングにおいて補間近似しない場合には、各々の式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合における各々の深さ方向の波数は、角周波数ωを伝搬速度ｃと式（Ｅ１）にて除したものである。以下、同様である。 For the two-dimensional Fourier transform R''(k _x ,0,k) computed in the case of methods (including method (1)) other than the normal migration (without steering in method (2)), Or, for R(k _x ,0,k) computed above in the normal migration, through an interpolation approximation (such as bi-linear interpolation using the nearest angular spectrum of frequency coordinates): , wavenumber matching represented by equation (M4) or equation (M4′) is performed.

However, s=2 if the received signal is a reflected signal, and s=1 if it is a transmitted signal.
However, when interpolation approximation is not performed in the wave number matching of formulas (M4) and (M4'), the wave number in the depth direction indicated by the proviso in each formula is used. Each wavenumber in the depth direction is obtained by dividing the angular frequency ω by the propagation velocity c and the equation (E1). The same applies hereinafter.

さらに、関数（Ｍ４''）を用いて、次の式（Ｍ５）又は式（Ｍ５'）が求められる。

Wavenumber matching is performed in this way to obtain the following function (M4'').

Furthermore, the following formula (M5) or formula (M5') is obtained using the function (M4'').

式（Ｍ６）及び式（Ｍ６'）の計算における２次元逆フーリエ変換は、高速２次元逆フーリエ変換（ＩＦＦＴ）を施せば良く、特殊な高速２次元逆フーリエ変換を使用することもできるが、一般的（popular）な方法としては、式（Ｍ６）及び式（Ｍ６'）の各々において、まず、信号帯域内の一方の波数ｋ_ｘに対し、他方の波数（Ｅ３）に関する高速逆フーリエ変換を行い、その上で、生成された空間座標ｙの各座標に対し、（信号帯域内の）波数ｋ_ｘに関する高速逆フーリエ変換を行えば良い（２次元のイメージ信号の各々を式（Ｍ６）又は式（Ｍ６'）に従って計算するよりは高速である）。 _Image A signal f(x,y) is generated.

The two-dimensional inverse Fourier transform in the calculation of formulas (M6) and (M6') may be performed by fast two-dimensional inverse Fourier transform (IFFT), and a special fast two-dimensional inverse Fourier transform may be used. As a popular method, in each of the equations (M6) and (M6′), first, one wavenumber _kx in the signal band is subjected to an inverse fast Fourier transform with respect to the other wavenumber (E3) and then perform an inverse fast Fourier transform with respect to the wavenumber kx (within the signal band) for each coordinate of the generated spatial coordinate _y (each of the two-dimensional image signals is expressed by the equation (M6) or faster than calculating according to equation (M6')).

Non-Patent Document 12 does not disclose the formula (M6) using y _S in the formula, but discloses that calculation is performed using y instead of y _S , and the coordinates are corrected after the calculation. . Coordinates are corrected by approximation or without approximation by multiplying with time based on complex exponential function multiplication, which is a past invention of the present inventor. Equation (M6) can also be used when the deflection angle is 0°.

非特許文献１２には、式中においてｙ_Ｓを用いる式（Ｍ７）は開示されていない。式（Ｍ７）は、偏向角度が０°のときにも使用できる。方法（１）～（６）と同様、スペクトルの周波数ｋ成分を足し合わせた上で、横方向の波数ｋ_ｘに関する逆フーリエ変換（ＩＦＦＴ）を行い、１回の逆フーリエ変換で計算ができ、計算が高速である。 In the apparatus of the present invention, wave number matching is performed together with a two-dimensional inverse Fourier transform, or together with a depth direction inverse Fourier transform, without interpolation approximation. In other words, the two-dimensional Fourier transform R''(k _x ,0,k) calculated in the case of methods (including method (1)) other than the normal migration method (when no steering in method (2)) is , or for the above R(k _x ,0,k) calculated in the normal migration method, first, in-band For each kx, integration with respect to _k is performed, and wavenumber matching of the wavenumber (E3) in the depth direction and inverse Fourier transform (IFFT) are performed simultaneously (step S34). Performs an inverse fast Fourier transform.

Non-Patent Document 12 does not disclose the formula (M7) using y _S in the formula. Equation (M7) can also be used when the deflection angle is 0°. As in methods (1) to (6), after summing the frequency _k components of the spectrum, inverse Fourier transform (IFFT) is performed on the horizontal wavenumber kx, and calculation can be performed with one inverse Fourier transform. Computation is fast.

纏めると、式（Ｍ６）、式（Ｍ７）、又は、式（Ｍ７'）を使用して、新しい処理を行うことができ、補間近似による誤差を除き、高速に、イメージ信号f(x,y)を生成できる。 Furthermore, when different from normal migration (processing corresponding to method (2) without steering), in the process of calculating equations (M6) and (M7), correction of the position in the lateral (x) direction It can be performed. For example, when the polarized plane wave of method (1) is transmitted, first, in step S34, the wave number (E3) is calculated as described above, and in step S35, each result is obtained for position correction. The function (M8) is multiplied by the complex exponential function (M9), and then, in step S36, an inverse fast Fourier transform (IFFT) is performed on the transverse wavenumber _kx . Alternatively, it may be calculated by multiplying the complex exponential function of the inverse Fourier transform with respect to the transverse wavenumber _{kx by} the formula (M9), or a dedicated fast inverse Fourier transform may be performed. Equation (M9) can also be used when the deflection angle is 0°. As described above, the image signal f(x,y) is generated in step S37.

In summary, using equation (M6), equation (M7), or equation (M7′), new processing can be performed to remove errors due to interpolation approximation, and at high speed, image signal f(x,y ) can be generated.

即ち、波数マッチングにおいて、補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合における深さ方向の波数は、角周波数ωを伝搬速度（Ｅ１）にて除したものである。
この式中に表される深さ方向の波数の式は、方法（１）における式（１３）に類似しているが、方法（１）において、式（１３）～式（１５）中のkx-ksinθの-ksinθを用いずに（零として）計算した場合に同じ結果を得る場合には、方法（６）において式（Ｍ９）を乗じて逆フーリエ変換を行う場合と同様に、段落０２０４に記載の処理を行う前に、式（１６）に式（Ｍ９）を乗じれば良い。但し、方法（６）で実現される平面波の偏向は、あくまで、近似計算の下で実現されるものであり、従って、方法（６）において式（Ｎ４）と式（Ｍ７）を用いた場合には全くに近似処理無しにて高精度化されるが、方法（１）において式（Ｍ９）を用いると精度は低下する。また、最後の逆フーリエ変換を２次元高速逆フーリエ変換で実施する場合（後述の通り、３次元の場合には３次元高速逆フーリエ変換）には、それらの処理を行うと、計算速度は、方法（６）では高速化されるが、方法（１）では遅くなる（段落０２０４に記載の処理は高速である）。
また、これらの改変された方法（１）や方法（６）の各々において、波数マッチングにおいて補間近似処理を行う場合において、式（Ｍ９）と（Ｎ４）とを用いて同じ結果を得る場合には、補間近似の式は対応して変化する（段落０３５４の（Ａ）と（Ｂ）の各々に記載）。
また、これらの改変された方法（１）や方法（６）の各々において、波数マッチングにおいて補間近似処理を行う場合において、上記の式（１１）と（Ｍ３）とを用いる場合（偏向角度データθを用いる）と、式（１１）と（Ｍ３）とを用いない場合（偏向角度θを零とする）とにおいて同結果を得る場合も、補間近似の式は対応して変化する（用いない場合を段落０３５４の（Ａ'）と（Ｂ'）の各々に記載）。
平面波送波を用いる場合のビームフォーミングは、本願明細書に記載の通り、様々なビームフォーミングに応用されるが、それらの応用において本段落に記載の処理が代わりに使用されることもある。注意すべきこととして、方法（６）を応用して送信フォーカシングされたもの等の任意の送信ビームフォーミングが行われたものに対して受信ダイナミックフォーカシングを行う場合には、後述の通り、角周波数ωと換算伝搬速度（Ｅ１）とを用いて表される波数（式（Ｍ１３））を用いて表される（Ｍ３''）を式（Ｍ３）の代わりに用いるため、補間近似式中の-ksinθ_tの波数kには、代わりに式（Ｍ１３）を使用する必要がある。
方法（１）と方法（６）との各々において、本段落に記載の方法が組み合わされて実施されることもある。例えば、方法（１）にて補間近似を行うJ.-y. Luらの方法（段落０１９７参照、式は段落０３５４の（Ｃ'）に記載）と同様に、方法（６）において、波数マッチングを全て補間近似を通じて行い、最初の２次元フーリエ変換と最後の２次元逆フーリエ変換を高速２次元フーリエ変換で実施することが可能である（段落０３５４の（Ｄ'）に記載。後述の通り、３次元の場合には３次元高速フーリエ変換）。これらの各々において、式（１１）を用いる場合（偏向角度データθを用いる）と式（Ｍ３）を用いる場合もある（段落０３５４の（Ｃ）と（Ｄ）の各々に記載）。無論、偏向しない場合にも使用できる。
上記の通り、方法（１）と方法（６）とを基礎とする平面波送波は、様々なビームフォーミングに応用される。 In addition, if the same result is obtained without multiplying the formula (M9) and without the approximation process, the following formula (N4) instead of the formula (M4) is used to calculate the formula (M6) or the formula (M7). good.

That is, in wavenumber matching, when interpolation approximation is not performed, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by the speed (E1).
The depth-direction wavenumber equation expressed in this equation is similar to equation (13) in method (1), but in method (1), kx If the same result is obtained when calculating without using -ksin θ of -ksin θ (as zero), in the same way as in the case of multiplying formula (M9) and performing inverse Fourier transform in method (6), paragraph 0204 Equation (16) may be multiplied by Equation (M9) before performing the described processing. However, the deflection of the plane wave realized by method (6) is only realized under approximation calculation. is highly accurate without any approximation processing, but the accuracy decreases when using equation (M9) in method (1). Also, when the final inverse Fourier transform is performed by a two-dimensional inverse fast Fourier transform (as described later, in the case of three dimensions, a three-dimensional inverse fast Fourier transform), the calculation speed is Method (6) speeds up, but method (1) slows down (the process described in paragraph 0204 is fast).
Also, in each of these modified methods (1) and (6), when interpolation approximation processing is performed in wave number matching, if the same result is obtained using equations (M9) and (N4) , the formula for the interpolation approximation changes correspondingly (described in paragraph 0354 (A) and (B), respectively).
Further, in each of these modified methods (1) and (6), when performing interpolation approximation processing in wave number matching, when using the above equations (11) and (M3) (deflection angle data θ ) and without using equations (11) and (M3) (with the deflection angle θ set to zero), the equation for interpolation approximation changes correspondingly (without using in each of (A') and (B') of paragraph 0354).
Beamforming with plane wave transmission has various beamforming applications, as described herein, in which the processing described in this paragraph may be used instead. It should be noted that when receiving dynamic focusing is performed on an object subjected to arbitrary transmission beamforming such as one subjected to transmission focusing by applying method (6), as will be described later, the angular frequency ω and the converted propagation velocity (E1) is used instead of the wavenumber (formula (M13)) (M3'') expressed by the formula (M3). For wavenumber k of _t , equation (M13) should be used instead.
In each of method (1) and method (6), the methods described in this paragraph may be combined and implemented. For example, in method (6), wavenumber matching are all performed through interpolation approximation, and the first two-dimensional Fourier transform and the final two-dimensional inverse Fourier transform can be performed with the fast two-dimensional Fourier transform (described in paragraph 0354 (D'). As described below, 3D Fast Fourier Transform for the 3D case). In each of these, equation (11) may be used (using the deflection angle data θ) and equation (M3) may be used (described in paragraph 0354 (C) and (D), respectively). Of course, it can also be used when it is not deflected.
As described above, plane wave transmission based on methods (1) and (6) has various beamforming applications.

Here, using the migration method, it was mainly described that the beamforming at the time of plane wave transmission with and without steering in method (1) is performed at high speed without interpolation approximation in wave number matching, but other methods of the present invention Method (2) (monostatic aperture synthesis with steering), Method (3) (multistatic with or without steering), Method (4) (transmit fixed focus with or without steering), and , method (5) (beamforming in a polar coordinate system or any orthogonal curvilinear coordinate system), respectively, can be performed similarly. Even if the deflection angles are different in transmission and reception, it can be processed in the same way. Apodization may be done as well.

A three-dimensional case can be similarly processed. When the received signal obtained using the two-dimensional aperture element array is represented by r(x, y, z, t), the received signal at the aperture element array position (y=0) is r(x, y= 0,z,t).

ここで、ｋ＝ω／ｃである。 First, as shown in equation (M'2), the received signal is subjected to three-dimensional Fourier transform with respect to time t, horizontal direction x, and elevation direction z (three-dimensional fast Fourier transform is preferable).

where k=ω/c.

In general, the received signal is subjected to fast Fourier transform (FFT) with respect to time t at each position coordinate (x, 0, z) to obtain the spectrum R (x, 0, z, k) of the analytic signal. be done. Then, at each frequency coordinate within the band k, a fast Fourier transform is performed in the horizontal direction x and the elevation direction z to obtain R(k _x ,0,k _z ,k) (each of the three-dimensional spectra is faster than calculating according to equation (M'2)).

When the plane wave is not transmitted with deflection, the above calculation is sufficient. must be trimmed by multiplying the above R'(x,0,z,k) after the fast Fourier transform for time t by a complex exponential function (M'3) (Fast Fourier transform for time t The transform and complex exponential multiplication operations can be computed directly at once, or a dedicated Fast Fourier Transform capable of such computation is useful).

Then, a fast Fourier transform (FFT) is applied to the received signal in the horizontal direction x. Let the result be R''(k _x ,0,z,k). By the way, even if the program is programmed to perform trimming, it is possible to process the case where a plane wave is transmitted without being deflected (steering angle 0°).

Next, wave number matching (or mapping) is performed. In the case of beamforming methods (1) to (5) other than normal migration (method (2) without steering), the propagation velocity c and the coordinate system (x _s , y _s , z _s ) are The propagation velocity (E'1) and the coordinate system (E'2) for each beam forming are read and processed.

但し、受信信号が反射信号の場合には、s = 2であり、透過信号の場合には、s = 1である。
但し、式（Ｍ'４）と式（Ｍ'４'）の波数マッチングにおいて補間近似しない場合には、各々の式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合における各々の深さ方向の波数は、角周波数ωを伝搬速度ｃと式（Ｅ'１）にて除したものである。以下、同様である。 For the three-dimensional Fourier transform R''(k _x ,0,z,k) computed in the case of methods (including method (1)) other than the normal migration (without steering in method (2)) or for R(k _x ,0,z,k) above computed in the normal migration, an interpolation approximation (using the angular spectrum nearest the frequency coordinate, bi-linear interpolation, etc.), the wave number matching represented by equation (M′4) or equation (M′4′), respectively, is performed.

However, s=2 if the received signal is a reflected signal, and s=1 if it is a transmitted signal.
However, when interpolation approximation is not performed in the wavenumber matching of formulas (M'4) and (M'4'), the wavenumbers in the depth direction indicated by the proviso in each formula are used, but interpolation approximation is obtained by dividing the angular frequency ω by the propagation velocity c and the equation (E′1). The same applies hereinafter.

さらに、関数（Ｍ'４''）を用いて、次の式（Ｍ'５）又は式（Ｍ'５'）が求められる。

Wavenumber matching is performed in this way to obtain the following function (M'4'').

Furthermore, using the function (M'4''), the following formula (M'5) or formula (M'5') is obtained.

For equation (M'5) or (M'5'), the wavenumbers k _x and k _z , for the three-dimensional case, as expressed by equation (M'6) or (M'6'). An image signal f(x,y,z) is generated by performing a three-dimensional inverse Fourier transform on the wavenumber (E'3).

The three-dimensional inverse Fourier transform in the calculation of the formulas (M'6) and (M'6') may be performed by performing a fast three-dimensional inverse Fourier transform (IFFT), and using a special fast three-dimensional inverse Fourier transform. However, as a popular method, in each of the equations (M'6) and (M'6'), first, for the two wavenumbers k _x and k _z in the signal band, Perform an inverse fast Fourier transform with respect to the other wavenumber (E′3), and then for each coordinate of the generated spatial coordinate y, with respect to wavenumbers k _x and k _z (within the signal band) (which is faster than calculating each of the three-dimensional image signals according to equation (M'6) or (M'6')).

In the apparatus of the present invention, wavenumber matching is performed together with a three-dimensional inverse Fourier transform or together with a depthwise inverse Fourier transform without interpolation approximation. That is, the three-dimensional Fourier transform R''(k _x ,0, k _z , k), or for the above R (k _x ,0, k _z ,k) calculated in the normal migration method, expressed by formula (M'7) or formula (M'7') First, for each (k _x , k _z ) in the band, the integration with respect to k is performed, and wavenumber matching and inverse Fourier transform (IFFT) of the wavenumber in the depth direction (E′3) are performed as shown in simultaneously, followed by an inverse fast Fourier transform in the lateral (x) and elevation (z) directions.

Non-Patent Document 12 does not disclose formulas (M'6) and (M'7) using _yS in the formulas. Both equations can also be used when the deflection angle is 0°. As in methods (1) to (6), after summing the frequency k components of the spectrum, inverse Fourier transform (IFFT) is performed on the wavenumbers k _x and k _z in the horizontal and elevation directions, and the inverse of 1 degree is obtained. It can be calculated by Fourier transform, and the calculation is fast.

纏めると、式（Ｍ'６）、式（Ｍ'６'）、式（Ｍ'７）、又は、式（Ｍ'７'）を使用して、新しい処理を行うことができ、補間近似による誤差を除き、高速に、イメージ信号f(x,y,z)を生成できる。 Furthermore, when different from normal migration (processing corresponding to method (2) without steering), in the process of calculating equations (M'6) and (M'7), horizontal (x) direction and the position in the elevation direction (z) can be corrected. For example, when the polarized plane wave of method (1) is transmitted, first, as described above, the calculation regarding the wave number (E'3) is performed, and for position correction, the function (M'8) obtained as a result of each is multiplied by a complex exponential function, followed by an inverse fast Fourier transform (IFFT) with respect to the transverse and elevational _{wavenumbers kx} and kz.

In summary, the new process can be performed using Eq. An image signal f(x, y, z) can be generated at high speed by removing errors.

即ち、波数マッチングにおいて、補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合における深さ方向の波数は、角周波数ωを伝搬速度（Ｅ'１）にて除したものである。
この式中に表される深さ方向の波数の式は、方法（１）における式（Ｃ２２）に類似しているが、方法（１）において、式（Ｃ２２）と式（Ｃ２３）中のkx-ksinθcosφとkz-ksinθsinφの-ksinθcosφと-ksinθsinφを用いずに（零として）計算した場合に同じ結果を得る場合には、方法（６）において２次元の場合の式（Ｍ９）に該当する複素指数関数を乗じて逆フーリエ変換を行う場合と同様に、段落０２０７に記載の処理を行う中で乗じれば良い。但し、方法（６）で実現される平面波の偏向は、あくまで、近似計算の下で実現されるものであり、従って、２次元の場合と同様に、方法（６）において式（Ｎ'４）と式（Ｍ７）を用いた場合には全くに近似処理無しにて高精度化されるが、方法（１）において２次元の場合の式（Ｍ９）に該当する複素指数関数を用いると精度は低下する。また、最後の逆フーリエ変換を３次元高速逆フーリエ変換で実施する場合には、それらの処理を行うと、２次元の場合と同様に、計算速度は、方法（６）では高速化されるが、方法（１）では遅くなる（段落０２０４に記載の処理は高速である）。
また、これらの改変された方法（１）や方法（６）の各々において、波数マッチングにおいて補間近似処理を行う場合において、２次元の場合の式（Ｍ９）に該当する複素指数関数と式（Ｎ'４）とを用いて同じ結果を得る場合には、２次元の場合と同様に、補間近似の式は対応して変化する（段落０３５４の（Ａ）と（Ｂ）の各々に記載）。
また、これらの改変された方法（１）や方法（６）の各々において、波数マッチングにおいて補間近似処理を行う場合において、上記の式（Ｃ２１）と（Ｍ'３）とを用いる場合（偏向角度データθとφを用いる）と、式（Ｃ２１）と（Ｍ'３）とを用いない場合（全ての偏向角度θとφを零とする）とにおいて同結果を得る場合も、２次元の場合と同様に、補間近似の式は対応して変化する（用いない場合を段落０３５４の（Ａ'）と（Ｂ'）の各々に記載）。
平面波送波を用いる場合のビームフォーミングは、本願明細書に記載の通り、様々なビームフォーミングに応用されるが、それらの応用において本段落に記載の処理が代わりに使用されることもある。２次元の場合と同様にして注意すべきこととして、方法（６）を応用して送信フォーカシングされたもの等の任意の送信ビームフォーミングが行われたものに対して受信ダイナミックフォーカシングを行う場合には、後述の通り、角周波数ωと換算伝搬速度（Ｅ'１）とを用いて表される波数（式（Ｍ'１３））を用いて表される（Ｍ'３''）を式（Ｍ'３）の代わりに用いるため、補間近似式中の-ksinθ₁(cosφ₁x+sinφ₁z)の波数kには、代わりに式（Ｍ'１３）を使用する必要がある。
方法（１）と方法（６）との各々において、本段落に記載の方法が組み合わされて実施されることもある。例えば、方法（１）にて補間近似を行うJ.-y. Luらの方法（段落０１９７参照、式は段落０３５４の（Ｃ'）に記載）と同様に、方法（６）において、波数マッチングを全て補間近似を通じて行い、最初の３次元フーリエ変換と最後の３次元逆フーリエ変換を高速フーリエ変換で実施することが可能である（段落０３５４の（Ｄ'）に記載)。これらの各々において、式（Ｃ２１）を用いる場合（偏向角度データθを用いる）と式（Ｍ'３）を用いる場合もある（段落０３５４の（Ｃ'）と（Ｄ'）の各々に記載）。無論、偏向しない場合にも使用できる。
上記の通り、方法（１）と方法（６）とを基礎とする平面波送波は、様々なビームフォーミングに応用される。 In addition, if the same result is obtained without approximation processing without multiplying the complex exponential function corresponding to the two-dimensional expression (M9), the following expression (N'4) can be used instead of the expression (M'4). can be used to calculate formula (M6) or formula (M7).

That is, in wavenumber matching, when interpolation approximation is not performed, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by the velocity (E'1).
The depth-direction wavenumber equation expressed in this equation is similar to equation (C22) in method (1), but in method (1), equation (C22) and kx If the same result is obtained without using -ksinθcosφ and -ksinθsinφ of -ksinθcosφ and kz-ksinθsinφ (as zero), the complex As in the case of performing inverse Fourier transform by multiplying by an exponential function, the multiplication may be performed while performing the processing described in paragraph 0207. However, the deflection of the plane wave realized in method (6) is only realized under approximation calculation, and therefore, in method (6), equation (N'4) and (M7) are used, the accuracy is improved without any approximation, but if the complex exponential function corresponding to the two-dimensional case (M9) is used in method (1), the accuracy is descend. Also, when the final inverse Fourier transform is performed by a three-dimensional fast inverse Fourier transform, these processes speed up the computation in method (6) as in the two-dimensional case, but , method (1) is slow (the process described in paragraph 0204 is fast).
Further, in each of these modified methods (1) and (6), when interpolation approximation is performed in wave number matching, the complex exponential function corresponding to the two-dimensional case (M9) and the expression (N '4) to obtain the same result, as in the two-dimensional case, the formula for the interpolation approximation changes correspondingly (described in paragraphs 0354 (A) and (B), respectively).
Further, in each of these modified methods (1) and (6), when performing interpolation approximation processing in wavenumber matching, when using the above equations (C21) and (M'3) (deflection angle Data θ and φ are used) and equations (C21) and (M′3) are not used (all deflection angles θ and φ are set to zero). Similarly, the formula for interpolation approximation changes correspondingly (cases not used are described in (A') and (B') of paragraph 0354, respectively).
Beamforming with plane wave transmission has various beamforming applications, as described herein, in which the processing described in this paragraph may be used instead. As in the two-dimensional case, it should be noted that when receiving dynamic focusing is performed on an object subjected to arbitrary transmission beamforming, such as one subjected to transmission focusing by applying method (6), , As will be described later, (M′3″) represented using the wave number (equation (M′13)) represented using the angular frequency ω and the reduced propagation velocity (E′1) is expressed by the equation (M Since it is used instead of '3), it is necessary to use the equation (M'13) instead for the wave number k of -ksin θ ₁ (cosφ ₁ x+sin φ ₁ z) in the interpolation approximation.
In each of method (1) and method (6), the methods described in this paragraph may be combined and implemented. For example, in method (6), wavenumber matching can all be done through an interpolation approximation, and the first 3D Fourier transform and the final 3D inverse Fourier transform can be performed with the Fast Fourier Transform (described in paragraph 0354 (D')). In each of these, there are cases where formula (C21) is used (using deflection angle data θ) and formula (M'3) is sometimes used (described in each of (C') and (D') in paragraph 0354) . Of course, it can also be used when it is not deflected.
As described above, plane wave transmission based on methods (1) and (6) has various beamforming applications.

In this migration method, similarly to the method (2) and method (3), monostatic and multistatic aperture plane synthesis can be performed without performing interpolation approximation processing in wave number matching.

を用いて表される

を同様に用い、通常のマイグレーション処理の波数マッチングにおいて補間処理を要しない場合の式（Ｍ７'）において、

但し、受信信号が反射信号の場合には、s = 2であり、透過信号の場合には、s = 1である。
とすれば良い。 In the case of monostatic aperture plane synthesis, if the soft transmission deflection angle and reception angle are θt and θr, instead of Equation (M3), the ultrasonic angular frequency ω ₀ and propagation velocity c are used to express wave number

is represented using

is used in the same way, and in the equation (M7') when interpolation processing is not required in wavenumber matching for normal migration processing,

However, s=2 if the received signal is a reflected signal, and s=1 if it is a transmitted signal.
and should be.

Also, in the three-dimensional case, if the deflection angles of the transmission beam and the reception beam are expressed using (elevation angle, azimuth angle)=(θ _t , φ _t ) and (θ _r , φ _r ), , wave number matching is performed first in the horizontal direction by multiplying by a complex exponential function (equation (D41)) expressed using the carrier frequency ω of the ultrasonic signal, as in method ( ₂ ), and then in the depth y Regarding the direction, in order to provide resolution in the depth y direction, the complex exponential function (formula (D42)) excluding the horizontal matching process (formula (D41)) is multiplied, and at the same time, the complex exponential function (formula (D43 )). That is, the expression (D41) is used instead of the expression (M3′) for the two-dimensional case, and the product of the expression (D42) and the expression (D43) is used instead of the expression (M11).

Thus, the migration processing in the present invention corresponding to method (2) and method (3) based thereon is equivalent to method (2) and method (3).

但し、受信信号が反射信号の場合には、s = 2であり、透過信号の場合には、s = 1である。また、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを伝搬速度ｃにて除したものである。以下、同様である。
と表される式（Ｍ４''）を求め、（Ｍ４'''）の同kyを用いて表される式（Ｍ５'）の２次元逆フーリエ変換（式（Ｍ６'））を実施するか、又は、式（Ｍ４'）の代わりに、

但し、受信信号が反射信号の場合には、s = 2であり、透過信号の場合には、s = 1である。また、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを伝搬速度ｃにて除したものである。以下、同様である。
と表される補間処理を共なう波数マッチングを行って式（Ｍ４''）を求め、（Ｍ４''''）の同kyを用いて表される式（Ｍ５'）に、

を乗算したものの２次元逆フーリエ変換（式（Ｍ６'）に該当）を実施すれば良い。 In these cases as well, it is possible to perform fast inverse Fourier transform by performing interpolation processing in wave number matching, as in normal migration processing. 3), instead of equation (M4′), which is wavenumber matching with interpolation after the above processing using equation (M3′), etc., based on equation (M11),

However, if the received signal is a reflected signal, s=2, and if it is a transmitted signal, s=1. When interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by c. The same applies hereinafter.
and perform the two-dimensional inverse Fourier transform (formula (M6')) of formula (M5') expressed using the same ky of (M4''') , or instead of formula (M4′),

However, if the received signal is a reflected signal, s=2, and if it is a transmitted signal, s=1. When interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by c. The same applies hereinafter.
Equation (M4'') is obtained by performing wavenumber matching accompanied by interpolation processing represented by, and Equation (M5') represented by using the same ky of (M4'''') is:

is multiplied by two-dimensional inverse Fourier transform (corresponding to formula (M6')).

Also in this case, multi-static aperture synthesis is performed in exactly the same manner as when method (2) is applied to realize method (3), instead of method (2), monostatic aperture synthesis by this migration method is performed. It can be realized by applying the aperture synthesis method.

但し、受信信号が反射信号の場合には、s = 2であり、透過信号の場合には、s = 1である。また、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを伝搬速度ｃにて除したものである。以下、同様である。
と表される式（Ｍ'４''）を求め、（Ｍ'４'''）の同kyを用いて表される式（Ｍ'５'）の３次元逆フーリエ変換（式（Ｍ'６'））を実施するか、又は、式（Ｍ'４'）の代わりに、

但し、受信信号が反射信号の場合には、s = 2であり、透過信号の場合には、s = 1である。また、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを伝搬速度ｃにて除したものである。以下、同様である。
と表される補間処理を共なう波数マッチングを行って式（Ｍ'４''）を求め、（Ｍ'４''''）の同kyを用いて表される式（Ｍ'５'）に、

を乗算したものの３次元逆フーリエ変換（式（Ｍ'６'）に該当）を実施すれば良い。 A three-dimensional case can be similarly processed. That is, after the processing using the equation (M'3') in the case of three dimensions, instead of the equation (M'4'), which is wave number matching with interpolation processing, the equation (D42) and the equation ( D43) (corresponding to (M11) in two dimensions),

However, s=2 if the received signal is a reflected signal, and s=1 if it is a transmitted signal. When interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by c. The same applies hereinafter.
and the three-dimensional inverse Fourier transform of the formula (M'5') expressed using the same ky of (M'4''') (formula (M' 6′)) or instead of formula (M′4′)

However, s=2 if the received signal is a reflected signal, and s=1 if it is a transmitted signal. When interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by c. The same applies hereinafter.
Equation (M'4'') is obtained by performing wavenumber matching accompanied by interpolation processing represented as, and equation (M'5') represented using the same ky of (M'4'''') ) to

is multiplied by the three-dimensional inverse Fourier transform (corresponding to formula (M'6')).

In this case as well, multi-static aperture synthesis is performed in exactly the same manner as when method (2) is applied to realize method (3), instead of method (2), monostatic aperture plane synthesis by this migration method is performed. It can be realized by applying a synthesis method.

Based on these migration methods, all beamforming described in methods (2) and (3) can be performed as well.

In addition, using the migration processing (equation (M7), etc.) at the time of plane wave transmission corresponding to method (1), without performing interpolation approximation in wavenumber matching, an arbitrary beam such as a fixed focus beam of method (4) In addition, beamforming when transmitting arbitrary waves (including waves that have not been beamformed), superposition processing when transmitting multiple beams and waves, and processing when transmitting multiple beams and waves at the same time can be performed. . Multiple beamforming may be performed using a multi-directional aperture synthesis method, which is similarly fast. Moreover, it is not limited to them. In those cases, receive polarized dynamic focusing can be implemented for arbitrary transmit beamforming in combination with method (2) in the same manner as with method (1).

を用いて表される

を同様に用い、また、平面波の物理的な送信偏向角度がＡであるときに、ソフト的な送信偏向角度と受信角度の各々をθ(＝θt)とθrとすると、式（Ｍ３）の代わりに、

を同様に用い、式（Ｍ７）において、

とすれば良い。 When the physical transmission deflection angle of the focused beam is A, and the soft transmission deflection angle and reception angle are θ (= θt) and θr respectively, instead of equation (M3), angular frequency ω and Wavenumber expressed using reduced propagation velocity (E1)

is represented using

is used in the same way, and when the physical transmission deflection angle of the plane wave is A, and the software transmission deflection angle and reception angle are θ (= θt) and θr respectively, instead of equation (M3), to the

is similarly used, and in equation (M7),

and should be.

を用いて表される

を同様に用い、また、平面波の物理的な送信偏向角度が仰角Ａと方位角Ｂで表されるとき（いずれかの少なくとも１つの角度が零度である場合を含む）、ソフト的な送信偏向角度(仰角θ₁と方位角φ₁)にてステアリングして偏向角度(仰角θ₂,方位角φ₂)にてステアリングダイナミックフォーカシングを行う場合（いずれかの少なくとも１つの角度が零度である場合を含む）には、式（Ｍ'３）の代わりに、

を同様に用い、式（Ｍ'７）において、

とすれば良い。 Also, the three-dimensional case can be similarly processed. When the physical transmit deflection angles of the focused beam are represented by the elevation angle A and the azimuth angle B (including the case where at least one angle is zero degrees), the soft transmit deflection angles (elevation angles θ ₁ and When steering at the azimuth angle φ ₁ ) and performing steering dynamic focusing at the deflection angles (elevation angle θ ₂ , azimuth angle φ ₂ ) (including the case where at least one of the angles is zero degrees), the formula Wavenumber expressed using angular frequency ω and reduced propagation velocity (E'1) instead of (M'3)

is represented using

is similarly used, and when the physical transmission deflection angles of the plane wave are represented by the elevation angle A and the azimuth angle B (including the case where at least one of the angles is zero degrees), the soft transmission deflection angle When steering is performed at (elevation angle θ ₁ and azimuth angle φ ₁ ) and steering dynamic focusing is performed at deflection angles (elevation angle θ ₂ , azimuth angle φ ₂ ) (including the case where at least one angle is zero ), instead of the formula (M'3),

is similarly used, and in formula (M'7),

and should be.

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ１）にて除したものである。以下、同様である。
と表される式（Ｍ４''）を求め、（Ｍ４'''''）の同kyを用いて表される式（Ｍ５）の２次元逆フーリエ変換（式（Ｍ６））を実施するか、又は、式（Ｍ４）の代わりに、

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ１）にて除したものである。以下、同様である。
と表される補間処理を共なう波数マッチングを行って式（Ｍ４''）を求め、（Ｍ４''''''）の同kyを用いて表される式（Ｍ５）に、

を乗算したものの２次元逆フーリエ変換（式（Ｍ６）に該当）を実施すれば良い。 In this case as well, as in normal migration processing, it is possible to perform fast inverse Fourier transform by performing interpolation processing in wavenumber matching. Instead of (M4), based on formula (M11'''),

However, when interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by the velocity (E1). The same applies hereinafter.
and perform the two-dimensional inverse Fourier transform (formula (M6)) of formula (M5) expressed using the same ky of (M4'''') , or instead of formula (M4),

However, when interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used. It is divided by the velocity (E1). The same applies hereinafter.
Equation (M4'') is obtained by performing wavenumber matching accompanied by interpolation processing represented by, and Equation (M5) represented by using the same ky of (M4'''''') is

is multiplied by two-dimensional inverse Fourier transform (corresponding to formula (M6)).

を用いると、ソフト的な受信ステアリングを行った場合（θrが非零の場合）には、生成される偏向角度が式（Ｍ１０）を用いた時よりも大きく生成される誤差(偏向角度２０°を実現する際に、１、２°程度)を生じるが、結像は得られる。 In the above, if the equations (M3'') and (M3''') are interchanged and each is processed, the error that the imaging position shifts when soft transmission steering is performed (when θt is non-zero) is produces Further, in these processes, the wavenumber expressed using the ultrasonic angular frequency ω ₀ and the converted propagation velocity (E1) instead of the equation (M10), which is the wavenumber corresponding to the ultrasonic frequency

(M10) produces a larger error (deflection angle 20° is about 1 or 2 degrees), but imaging is obtained.

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ'１）にて除したものである。以下、同様である。
と表される式（Ｍ'４''）を求め、（Ｍ'４'''''）の同kyを用いて表される式（Ｍ'５）の３次元逆フーリエ変換（式（Ｍ'６））を実施するか、又は、式（Ｍ'４）の代わりに、

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ'１）にて除したものである。以下、同様である。
と表される補間処理を共なう波数マッチングを行って式（Ｍ'４''）を求め、（Ｍ'４''''''）の同kyを用いて表される式（Ｍ'５）に、

を乗算したものの２次元逆フーリエ変換（式（Ｍ'６）に該当）を実施すれば良い。 Also, three-dimensional images can be processed in the same way. That is, after the processing using the equation (M'3'') in the case of three dimensions, instead of the equation (M'4), which is wave number matching with interpolation processing, the equation (M'11''')On the basis of,

However, when interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction shown in the proviso in the formula is used. It is divided by the velocity (E'1). The same applies hereinafter.
and the three-dimensional inverse Fourier transform of the equation (M'5) expressed using the same ky of (M'4''''') (Equation (M '6)), or instead of formula (M'4),

However, when interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction shown in the proviso in the formula is used. It is divided by the velocity (E'1). The same applies hereinafter.
Equation (M'4'') is obtained by performing wave number matching accompanied by interpolation processing represented by, and equation (M' 5) to

is multiplied by two-dimensional inverse Fourier transform (corresponding to formula (M'6)).

を用いると、ソフト的な受信ステアリングを行った場合（偏向角度が非零の場合）には、生成される偏向角度が式（Ｍ'１０）を用いた時よりも大きく生成される誤差を生じるが、結像は得られる。 In the above, if the equations (M'3'') and (M'3''') are interchanged and each is processed, the result will be This causes an image position error. Further, in these processes, the wavenumber expressed using the ultrasonic angular frequency ω ₀ and the converted propagation velocity (E′1) instead of the equation (M′10), which is the wavenumber corresponding to the ultrasonic frequency

using (M'10) produces an error that produces a larger deflection angle than when using equation (M'10) in the case of soft receive steering (non-zero deflection angle). but imaging is obtained.

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用され、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ１）にて除したものである。以下、同様である。 Also, in this case, when using the migration method based on the formula (N4) described in paragraph 0316, when the physical transmission deflection angle of the focus beam is A, the soft transmission deflection angle and reception angle are θ (= θt) and θr respectively, the wave number (M13) expressed using the angular frequency ω and the converted propagation velocity (E1) is used instead of the equation (M3) (M3'') is similarly used, and when the physical transmission deflection angle of the plane wave is A, and the software transmission deflection angle and reception angle are θ (=θt) and θr respectively, the equation (M3) is (M3''') may be similarly used instead of , and the following formula (N4') may be similarly used instead of Formula (N4) in Formula (M6) or (M7).

However, when interpolation approximation is not performed in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used, and the wavenumber in the depth direction when interpolation approximation is performed is obtained by converting the angular frequency ω into the converted propagation velocity ( E1). The same applies hereinafter.

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用され、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ'１）にて除したものである。以下、同様である。 Also, the case of three dimensions (when using the migration method based on the formula (N'4) described in paragraph 0331) can be similarly processed. When the physical transmit deflection angles of the focused beam are represented by the elevation angle A and the azimuth angle B (including the case where at least _one angle is zero degrees), the soft transmit deflection angles (elevation angles θ1 and When steering at the azimuth angle φ ₁ ) and performing steering dynamic focusing at the deflection angles (elevation angle θ ₂ , azimuth angle φ ₂ ) (including the case where at least one of the angles is zero degrees), the formula Instead of (M'3), the equation (M'3'') expressed using the wave number (M'13) expressed using the angular frequency ω and the reduced propagation velocity (E'1) is similarly Also, when the physical transmission deflection angle of the plane wave is represented by the elevation angle A and the azimuth angle B (including the case where at least one angle is zero degrees), the soft transmission deflection angle (elevation angle θ ₁ and azimuth angle φ ₁ ) to perform steering dynamic focusing at deflection angles (elevation angle θ ₂ , azimuth angle φ ₂ ) (including the case where at least one angle is zero) , in place of formula (M'3), similarly use formula (M'3'''), and in formula (M'6) or (M'7), instead of formula (N'4), Equation (N'4') may be used similarly.

However, when interpolation approximation is not performed in wavenumber matching, the wavenumber in the depth direction indicated by the proviso in the formula is used, and the wavenumber in the depth direction when interpolation approximation is performed is obtained by converting the angular frequency ω into the converted propagation velocity ( E'1). The same applies hereinafter.

In this way, beamforming based on these migration methods can also be performed with or without interpolation approximation processing in wavenumber matching in the same manner as the methods described in paragraphs 0316 and 0331.

In addition, the processing described in paragraphs 0316 and 0331 with respect to method (1) may be combined with method (1) and method (2) (method (1) described in paragraphs 0235 to 0238 when receiving deflection dynamic focusing is performed. ) can be implemented in the same way. That is, in the two-dimensional case, if the same result is obtained when the deflection angle θ in the formula (F42) and the formula (F43) is set to zero, then in the method (6), the formula (M9) is multiplied and the inverse As in the case of Fourier transform, the formula corresponding to formula (16) may be multiplied by formula (M9) before the process described in paragraph 0204 is performed. In the three-dimensional case, if the same result is obtained when all the deflection angles θ and φ in the equations (G22) and (G23) are set to zero, then in the two-dimensional case in method (6) (M9) is multiplied by the complex exponential function corresponding to the inverse Fourier transform.
In these beamformings, as described in paragraphs 0316 and 0331, interpolation processing can be performed in wavenumber matching, or it can be performed without performing it. Others are as described in the same paragraph.

Based on these migration methods, all beamforming described in method (4) can be performed as well.

Also, as described in method (5), when transmission and reception are performed in an orthogonal coordinate system other than the Cartesian coordinate system such as the polar coordinate system, the equations (M6), (M6'), and ( M7), the Jacobian version of equation (M7′) can be computed to perform the beamforming described above to obtain the result directly in the Cartesian coordinate system. Also, in the case of three dimensions, the Jacobi operation is applied to the equations (M'6), (M'6'), (M'7), and (M'7') in exactly the same way, can be processed. In addition, all beamforming described in method (5) can be performed similarly.

波数マッチングにおいて補間近似を行う場合の深さ方向の波数k_yは、角周波数ωを伝搬速度cにて除したものであるが、補間近似しない場合に上記の如く処理できる。

波数マッチングにおいて補間近似を行う場合の深さ方向の波数k_yは、角周波数ωを伝搬速度cにて除したものであるが、補間近似しない場合に上記の如く処理できる。

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ１）にて除したものである。

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ１）にて除したものである。

波数マッチングにおいて補間近似を行う場合の深さ方向の波数k_yは、角周波数ωを伝搬速度cにて除したものであるが、補間近似しない場合に上記の如く処理できる。

波数マッチングにおいて補間近似を行う場合の深さ方向の波数k_yは、角周波数ωを伝搬速度cにて除したものであるが、本発明の１つである、補間近似しない場合には、方法（１）に従って、上記の如く処理できる。

但し、本発明の１つである、波数マッチングにおいて補間近似しない場合（方法（６）の１つ）には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ１）にて除したものである。

但し、波数マッチングにおいて補間近似しない場合には、式中の但し書きに表される深さ方向の波数が使用されるが、補間近似を行う場合の深さ方向の波数は、角周波数ωを換算伝搬速度（Ｅ１）にて除したものである。
これらにおいても、受信ビームフォーミングにおいて、方法（２）が併用されることがある。
最初の多次元フーリエ変換と多次元逆フーリエ変換を高速に実施することは重要であり、適切に各種の高速フーリエ変換アルゴリズムを使用できる。また、物理的にも、数学的にも、最初のフーリエ変換時に波数マッチングを行うこともできるし、最後の逆フーリエ変換時に波数マッチングを行うことも可能である。また、本明細書に記載されているビームフォーミング（方法（１）～方法（６）等）以外のものも全て、同様に、補間近似処理無し又はその近似処理を通じた処理により実現できる。補間近似処理を行う場合において高精度化するべく、サンプリング周波数を高くすることがあるが、補間近似処理を行わない場合に任意位置の信号を生成できるのとは異なり、フーリ変換のデータ数が増えることには注意が必要である。但し、補間近似処理を行わない場合も、適度にオーバーサンプリングを行い、信号を高ＳＮ比化して処理する状況を実現することは重要である。
この様な中で、マイグレーション処理に基づいた場合（方法（６））においても、方法（１）～（５）に記載のビームフォーミング（複数の異なるビーム又は波動の同時送信時における一括受信信号や各々の送信に対する受信信号の重ね合わせのビームフォーミングや、仮想源や仮想受信機を用いる場合等を含む）を、補間近似を行う場合と、行わない場合とにおいて、実施できる。 One of the objects of the present invention is to realize high-speed and high-precision beamforming based on the Fourier transform. Any processing that does not have an approximation can be used after being modified into a processing that includes various forms of interpolation approximation, and although the accuracy is reduced, it may be used at an increased speed. Whether wave number matching in at least one direction, two directions, or all three directions out of the three directions of the lateral direction, the elevation direction, and the depth direction is performed by approximation, or multiplication of a complex exponential function is used, etc. Modifications of the method can be used. More approximations can speed things up, but reduce accuracy. In the explanation so far, some have already been explained. In this paragraph, (A), (A'), (B), (B'), (C), (C'), ( Approximate interpolation formulas are shown for the eight cases D) and (D').
For example, the same process can be performed when steering a plane wave by the migration method of method (6). Similar to J.-y. Lu et al.'s method of processing (see paragraph 0197, corresponding to (C')), the computation is fastest when migration is used. However, the accuracy is the lowest among migrations. On the other hand, in the method of J.-y. Lu et al. (see paragraph 0197, corresponding to (C')), for example, if only horizontal wavenumber matching is performed before the Fourier transform, the accuracy is improved, but the calculation speed is (corresponding to (C)). Including the cases of (A), (A'), (B), and (B'), the interpolation processing (formula) for the two-dimensional case (paragraph 0316) is shown below (for the three-dimensional case (paragraph 0331 ) are similarly represented and abbreviated). Incidentally, (A), (A'), (C), and (C') are represented according to equations (7) and (8). Also, in (B') and (D'), the inverse Fourier transform in the horizontal direction is performed on kx' instead of kx.

The wave number k _y in the depth direction when interpolation approximation is performed in wave number matching is obtained by dividing the angular frequency ω by the propagation velocity c, but can be processed as described above when interpolation approximation is not performed.

The wave number k _y in the depth direction when interpolation approximation is performed in wave number matching is obtained by dividing the angular frequency ω by the propagation velocity c, but can be processed as described above when interpolation approximation is not performed.

However, when interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction shown in the proviso in the formula is used. It is divided by the speed (E1).

However, when interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction shown in the proviso in the formula is used. It is divided by the speed (E1).

The wave number k _y in the depth direction when interpolation approximation is performed in wave number matching is obtained by dividing the angular frequency ω by the propagation velocity c, but can be processed as described above when interpolation approximation is not performed.

The wave number _ky in the depth direction when interpolation approximation is performed in wave number matching is obtained by dividing the angular frequency ω by the propagation velocity c. According to (1), it can be processed as described above.

However, when interpolation approximation is not performed in wave number matching, which is one of the present invention (one of method (6)), the wave number in the depth direction expressed in the proviso in the formula is used, but interpolation approximation is obtained by dividing the angular frequency ω by the converted propagation velocity (E1).

However, when interpolation approximation is not used in wavenumber matching, the wavenumber in the depth direction shown in the proviso in the formula is used. It is divided by the speed (E1).
In these cases, the method (2) may also be used in reception beamforming.
A fast implementation of the initial multidimensional Fourier transform and the multidimensional inverse Fourier transform is important, and various fast Fourier transform algorithms can be used as appropriate. Physically and mathematically, wave number matching can be performed during the first Fourier transform, and wave number matching can be performed during the final inverse Fourier transform. In addition, all of the beamforming methods other than the beamforming described in this specification (Methods (1) to (6), etc.) can be similarly realized without interpolation approximation processing or with processing through such approximation processing. When performing interpolation approximation processing, the sampling frequency may be increased in order to improve accuracy. It is necessary to pay attention to this. However, even if interpolation approximation processing is not performed, it is important to implement a situation in which signals are processed with a high SN ratio by appropriately oversampling.
Under such circumstances, even in the case based on migration processing (method (6)), the beamforming described in methods (1) to (5) (batch reception signals at the time of simultaneous transmission of multiple different beams or waves, beamforming of the superposition of the received signals on each transmit, using virtual sources and virtual receivers, etc.) can be implemented with and without interpolation approximation.

Method (7): Others Although the above methods (1) to (6) have mainly been described for the case of using a one-dimensional array, the same applies to a two-dimensional array or a three-dimensional array as described above. , and in each case, the other one or two other axial directions may be similarly treated as for the transverse direction. These can be implemented in any Cartesian coordinate system, including the Cartesian Cartesian coordinate system. That is, simply extend the above methods (1) to (6) to higher dimensions. In addition, during the processing of methods (1) to (6) and method (7) described herein, horizontal or vertical DC components and low frequency components may be generated. It is preferable to zero-fill the spectrum before the inverse Fourier transform of . At the start of digital signal processing, analog processing or digital processing is performed as preprocessing to cut direct current, and the spectrum may be zero-padded in the angular spectrum.
In addition, when performing a digital Fourier transform, the distribution of the raw signal before processing and the image signal after processing in a finite spatial region is assumed to be periodic (that is, cyclic), so that the upper and lower boundaries of the region of interest etc. (mainly boundaries parallel to the surface of the physical aperture element array, boundaries running in the so-called horizontal and elevation directions, etc.) and other side and vertical boundaries, etc. (mainly the surfaces of the physical aperture element array It may be a problem that the unprocessed raw signal component and the processed image signal appear circulating around a boundary in a direction perpendicular to the .
In both the case of targeting reflected waves and the case of targeting transmitted waves, the signal intensity decreases as it propagates away from the transmission aperture plane due to phenomena such as diffusion, attenuation, scattering, and reflection. have such raw signals, the upper and lower boundaries of those regions of interest (mainly the boundaries parallel to the plane of the physical aperture element array and the so-called boundaries running in the lateral and elevational directions). etc.) may not be a problem, but it is a problem when the intensity of the raw signal is not small at the top or bottom. If the receive aperture is different from the transmit aperture for transmitted waves, the signal strength at the receive aperture position may become a problem. In such a case, a region of zero signal (strictly speaking, the distance between the top and the bottom) is placed in the region of interest of the raw signal in the direction (longitudinal direction) approaching from the top of the region of interest toward the transmit aperture plane. or add a region of zero signal (strictly, a region with the length of the top and bottom distance) in the direction away from the transmit aperture plane (longitudinal direction) from the bottom of the region of interest. In addition, after the region of interest is expanded in the direction away from the transmit aperture and the receive aperture (longitudinal direction) and processed, the image signal obtained in the added region is discarded, or the raw signal is still discontinuous. Therefore, instead, the raw signal is multiplied by a so-called window in the longitudinal direction and processed, and the image signal obtained near the bottom of the window may not be emphasized.
On the other hand, at other side surfaces and vertical boundaries (mainly, boundaries perpendicular to the surface of the physical aperture element array and boundaries running in the so-called axial direction), raw signals representing different spatial positions are used. The same phenomenon (artifact) is a problem in most cases because the image signal is synthesized and the raw signal is often strong. In such a case, a region of interest in the raw signal may have a direction away from the region of interest (width direction) from at least one of the two paired boundaries of the lateral and vertical boundaries of the region of interest. The image signal obtained in the area added after adding the area of zero signal (strictly speaking, the area with the larger width of the effective aperture width for transmission or reception), expanding the area of interest in the horizontal direction and processing it However, since the raw signal has a discontinuous state, it can be processed by multiplying the raw signal by a so-called window in the width direction, and the image signal obtained near the bottom of the window is not emphasized. be. In the three-dimensional case, the windows are multidimensional in lateral and elevational directions, so-called segregated or non-isolated windows. This is also a problem when steering is not performed, but when steering is performed, it is often necessary to do so.
In some cases it may be useful to perform those processes described above at the same time. When windows are used, they are multi-dimensional, so-called separable or non-isolated windows. In addition, if there is an image signal that appears cyclically in the image signal obtained without these processes, the region of the image signal may be excluded from the region of interest.

を持つ。その方法を用いて、円筒座標系や球座標系、他の如何なる直交座標系においても、グリーン関数を用いて演算可能である。
つまり、方法（１）～（７）の中で開示した方法又は数式（波数マッチングにおいて補間近似を行わない場合と行う場合）において、求める信号のスペクトルが、分母に式（ＧＲ１）又は（ＧＲ２）を持つ状況において演算すれば良い。方法（１）～（７）に記載の方法や数式は、その他の様々な方法やビームフォーミングに応用できる。
また、グリーン関数を用いるこれらの場合において、点源を考えることができることから、物理開口の前後に設定される仮想源（特許文献７や非特許文献８）を表すのに適している。その場合に、次段落にて述べる物理開口（素子）の実際の放射パターンに対する処理は重要である。 In addition, the methods disclosed in methods (1) to (7) can also be used in beamforming using other Fourier transforms in Non-Patent Document 9, etc., and the same effect can be obtained. .
For example, Section 2.4 of Non-Patent Document 9 discloses a method of calculating an arbitrary beam or wave using a general solution (Green's function) of the wave equation. There, spherical waves, cylindrical waves, and plane waves are treated as examples of analytical calculations. As a feature of using the Green's function, the desired signal in the frequency domain is

have. Using that method, it is possible to operate with Green's functions in cylindrical, spherical, or any other Cartesian coordinate system.
That is, in the methods or formulas disclosed in methods (1) to (7) (when interpolation approximation is not performed in wave number matching and when interpolation approximation is performed), the spectrum of the signal to be sought is expressed by formula (GR1) or (GR2) in the denominator. The calculation should be performed in the situation where The methods and formulas described in methods (1) to (7) can be applied to various other methods and beamforming.
In addition, in these cases using the Green's function, point sources can be considered, which makes them suitable for representing virtual sources (Patent Document 7 and Non-Patent Document 8) set before and after the physical aperture. In that case, the treatment of the actual radiation pattern of the physical apertures (elements) described in the next paragraph is important.

In addition, the methods (1) to (7) can be applied with the computation taking into account the radiation pattern of the aperture (element) described in Section 3.2 of Non-Patent Document 9, for example. In that case, signal processing may be performed by appropriately correcting the signal intensity by physical or soft apodization. As many described in this specification, for example, ISAR (sometimes applying the movement of the observed object), nonlinear processing, adaptive beamforming (Non-Patent Document 10), etc., and various other processing In some cases, high resolution (especially in the direction orthogonal to the propagation direction) and high contrast by suppressing side lobes are achieved. A coherent factor or the like described in Non-Patent Document 11 or the like can also be applied. In addition, it is not limited to them, and apodization (a complex signal may also be used as a delay) is appropriately performed. Note that the apodization may be variable not only in the propagation direction but also in the scanning direction.

Methods (1)-(7) can also be used with transmit or receive apertures at different locations and with various other beamformings. For example, Non-Patent Document 9 presents a wealth of examples. For example, there are descriptions of Geophysical Imaging (for example, pay attention to the forms of Equations 7.9 to 7.12) and so-called X-ray CT (Computed Tomography) in Section 7.3. This method (1) to (7) can also be used for Also worth noting is Fig. 7.3 and Eqs. In these, processing including wavenumber matching can be performed without performing interpolation approximation processing (there are cases where interpolation approximation processing is performed).

In addition, the methods (1) to (7) are applied to a predetermined biplane, multiple planes, planes or tomographic planes extending in any desired direction (they are not necessarily flat such as planes or tomographic planes, but curved surfaces). It has the characteristic of being able to selectively and directly generate an image signal on a line rather than on a plane (a line is a straight line or a curve). For example, while images can be presented based on three-dimensional or two-dimensional image signals, images may be presented based on those image signals, or those images may be presented independently. It is possible. Also, in signal processing, these image signals and images are sometimes presented through interpolation approximation. Moreover, in them, measurement data such as displacement, strain, and temperature measured based on image signals and images may be displayed alone or may be displayed superimposed on those images.

As described multiple times herein, apodization can be determined and performed in a variety of ways. There are various adaptive beamforming, minimum variance beamforming, Capon method, etc. described in Non-Patent Document 10 and the like. In these, when regularizing the covariance matrix, the parameter for adjusting the degree of regularization (regularization parameter) can be appropriately determined for each position based on the SN ratio of the signal at each position. , can be spatially variant and processed. Alternatively, instead of using the identity matrix for the regularization operator, it is also possible to use other positive definite operators such as the gradient operator, the Laplacian, and so on. It is possible to increase the resolution of the image signal (especially in the direction orthogonal to the propagation direction), and it is also possible to increase the contrast by suppressing side lobes. Also in independent component analysis (independent signal separation), it is effective to regularize the covariance matrix in the same manner. These regularization methods have not been previously disclosed. Alternatively, both processes are stabilized by lowering the rank based on singular value decomposition or eigenvalue decomposition. These processes are also effective in other beamforming methods. As described elsewhere, MIMO (Multiple-input and Multiple-output: a wireless communication technology that combines multiple antennas on the transmitting side and the receiving side to widen the band of transmitted and received signals) and SIMO (Single-input and Multiple-output) : a wireless communication technology in which one antenna is used on the transmitting side and a plurality of antennas are used on the receiving side to widen the band of the received signal). Further, the inventor of the present application prefers to use absolute value detection and power-law detection in addition to envelope detection. Absolute value detection and power detection are effective in visualizing rippling of waves. If the absolute value is taken or raised to a power, especially if the order is high, it will have a high frequency component. can also be regarded as biased detection). Non-Patent Document 10 describes various other adaptive beamforming, and this specification also describes MUSIC (Multiple Signal Classification: wireless communication technology using eigenvalues or eigenvectors of correlation matrix of received signals), etc. Various processes are described. Also, effective processing is not limited to that, and there are various other processing. Such various treatments may be performed before, during, or after beamforming, and may be performed at the level of apodization. In that case, it is extremely effective to perform spatio-temporal alignment based on correlation processing before processing (details will be described later).

In addition, in the present invention, integration processing (calculation) may be performed particularly in the time direction (distance direction) in which the rf data is acquired to improve the signal-to-noise ratio of the signal. This integral processing is also performed by analog processing (so-called integrator) or digital processing (integrator or integral calculation).

Also, in the above description, the calculation of the product with the weight value, which is easy to perform with a small amount of calculation for the apodization process, has been described. Convolution may be performed based on the dual relationship between the product in and the convolution integral. Appropriate apodization is possible at each depth position or at every equal distance from each aperture element.

By superimposing the multidirectional polarized beams and plane waves generated by the apparatus according to the present embodiment using methods (1) to (6), the above laterally modulated signal (image signal) and laterally broadband In some cases, a signal (image signal) having a high resolution in the horizontal direction is generated. As with single transmissions, physical or soft steering or both may be performed on each, or they may all be subject to the same steering. Although reception beamforming is mainly performed by software, if necessary, reception beamforming using reception delay or reception apodization may be physically performed singly or in combination. On the other hand, although transmission beamforming is mainly described as being physically implemented, it is possible to obtain a high-speed frame rate by, for example, transmitting a plane wave or transmitting a plurality of beams or waves. In order to perform surface synthesis, transmission may be performed for each element (it is also possible to perform multi-directional aperture synthesis. Plane waves, cylindrical waves, spherical waves, etc. are transmitted by coding, and decoded to perform aperture synthesis. can also be done). In terms of software, it is possible to conceive and implement transmission and reception in reverse, as described above. A plane wave reaches a deeper position than a focused beam (an echo can also be obtained from a deeper position), but the SN ratio as a wave or beam for the purpose of measuring displacement is comparable. and low. In the first place, the lateral resolution is also relatively low. On the other hand, when plane waves in multiple directions are superimposed, substantially the same horizontal resolution is obtained regardless of the position. On the other hand, when using a focus beam, it is effective to cross beams in multiple directions at the focus position. It will be done. In any case, high-speed beamforming can be achieved with respect to the received signal when multiple waves or beams are transmitted simultaneously, or the superposition of the received signals transmitted at different times even if the target is in the same phase. . Also, a plurality of waves with different carrier frequencies may be used to generate a signal (image signal) that is axially broadened in band and axially high in resolution. In these cases, the bands may be widened so that the spectra overlap to achieve high resolution. These multiple beams may also be generated in parallel at the same time, and may be generated at different times even though the measurement targets are in the same time phase. Multi-directional waves may also be generated by the multi-directional aperture synthesis described above.
Incidentally, when the deflection angle is increased, the imaging positions of the reflector and the strong scatterer may be spatially displaced. For example, when the deflection angle is changed in small steps (for example, by 1°) up to ±45°, a plane wave is transmitted with respect to the front of the aperture element, and beam-formed signals are superimposed to form an aperture in the front direction. The case of surface synthesis can be simulated, and a band corresponding to a deflection angle of ±45° in the lateral direction cannot be obtained (signals are canceled during superposition). It is natural to consider the case where beam forming in the front direction is decomposed into plane waves as an angular spectrum. In the case of lateral band broadening by overlapping in space-time or frequency space, not only in the case of polarized plane wave transmission, but also in the case of other polarized beamforming, the spectrum must not overlap in frequency space. Although it is necessary to superimpose them, the cause of the spatial deviation of the imaging positions can be found there. In other words, as phase aberration, spatial inhomogeneity of the sound velocity of the medium (depending on temperature and pressure) and the directivity of the aperture, especially during steering with a large steering angle, cause errors in focusing and control of the propagation direction. Therefore, when superimposing signals with a large deflection angle, at least one of the time of transmission during beamforming, the received signal before reception beamforming, during reception beamforming, and after beamforming, It may be necessary to correct the position of the signal (phase aberration correction). When super-resolution is performed by spectral processing or nonlinear processing of these signals, when superimposition processing (spectral processing or nonlinear processing of superposition, or superposition of spectral processing or nonlinear processing, etc.) is used together, Such position corrections can be important, as the effects of positional deviations on results (positional errors, etc.) can be significant. In addition, the frequency dependence of the converted propagation velocity and the like are also factors that cause misalignment, and the same problem may occur when signals of different frequencies are superimposed, and may be processed in the same way. In addition, waveform distortion may occur due to frequency-dependent phase delays and phase changes in devices (for example, phase changes in sensors and phase delays in amplifiers in circuits). It may be corrected to shape the waveform (it is effective to perform analog-like shifting of the digital signal that rotates the phase of each signal component by multiplying it by a complex exponential function frequency by frequency in the frequency domain). Obtaining the phase change that occurs in the device in advance to generate a wave having a desired waveform, or appropriately electrically driving the sensor with the goal of generating a desired waveform at the observation position or reception position. There is also As a result, the observation of the observation target can be made with high accuracy (various observations using phase, observations of frequency-dependent propagation velocity, scattering, attenuation, etc.). Only the amount that occurs at the receiving device may be corrected. Other misregistrations mentioned above may also be processed in the same way. Position correction is also described in paragraph 0371 and the like. Various signal processing techniques for motion compensation and phase aberration correction can be used. Further, the signal strength correction is also described in paragraph 0689 and the like. Regarding the beamforming component and the speckle component of the weakly scattered signal, unlike those deterministic signals, the virtual sound source and virtual receiver assuming a scattering object invented by the inventor of the present application (Patent Document 7, Non-Patent Document 8), it can be used for imaging and displacement measurement without performing position correction processing (for example, it is effective to combine plane wave transmission and Gaussian apodization for displacement measurement). , a power-type apodization such as a square function is effective during focusing, and the latter has a high resolution even in single beamforming).

In addition, multi-focusing (not limited to normal multi-focusing in which transmission focuses are formed at multiple locations in the beam propagation direction, including lateral direction, including taking transmission focus at arbitrary multiple positions) A process of performing reception beamforming by transmitting a plurality of waves focused on a position may be performed. Further, processing for generating reception beams at a plurality of positions and reception beams in a plurality of directions may be performed for one transmission beam. Also, beamforming based on transmission in multiple directions with different steering angles may be performed. Others include beamforming at distant locations with little interference, and beamforming in each divided portion within a frame on the basis of transmitting and receiving, and multiple beamforming thereof. may be performed by performing parallel beamforming, such as by processing in parallel, each superimposing the results at each location in the region of interest where multiple beamforming results were obtained. Method (4) itself has the feature that it is possible to process received signals of interfering waves as long as the waves propagate within the region of interest, and by taking advantage of this, a high frame rate can be achieved. (Method (4) allows receive beamforming for any transmit beam or wave, single or multiple transmissions). Also, beam forming may be performed using appropriate signal components after performing various signal separation processes. Depending on the degree of interference, waves outside the region of interest may be processed without being removed.

In addition, the aperture for transmitting or receiving waves may be a dedicated aperture for each, but it may also serve as both. Waves may also be received, and the beamforming results generated may also be superimposed, including when processed in parallel. In summary, the above superposition can be performed at the same time, or at the same time when the state of the target (communication target) or the observed target is the same or substantially the same, or at a different time, or at a different time. There may be one or more beamformings or transmissions or receptions at each aperture in a phase, and similarly one or more beamformings or transmissions or receptions at one combination of apertures. It is sometimes done. Similarly, each of multiple combinations of apertures may perform one or more beamforming or transmission or reception. In addition, when multiple results of beamforming or transmission or reception are obtained in them, new data may be generated through computation of their superposition.

Since the superposition process is a linear process, in the calculation process of the above methods (1) to (6), it is possible to superimpose a plurality of complex spectral signals having the same frequency in the frequency domain, in which case In order to do so, it suffices to inverse Fourier transform the superimposed complex spectra at once, and when superimposing as described above in the spatial domain after generating a plurality of beamformed waves, inverse Fourier transforms are performed as many times as the number of waves to be superimposed. You can complete the process faster than you need to. Incoming waves, such as, but not limited to, superimposed angular spectra may be processed, for example, in a single direction or in multiple directions. As for the processing itself, it is preferable to superimpose a plurality of waves in the spatial domain, perform Fourier transform, and obtain the superimposed angular spectrum (there is an effect that only one Fourier transform is required). In some cases, it is possible to confirm the position of the target object.

Also, in the above, when transmitting a plurality of beamformed waves (at least, a predetermined transmission delay is applied, and a transmission apodization may be applied) (methods (1) to (6), In methods (2) and (3), which excludes aperture synthesis, the first excited aperture element in the effective aperture array for each wave to be transmitted (the first excited aperture element are excited at the same time), and the respective received signals are stored in a memory or storage device (storage medium) in an overlapping state. (Parallel processing may be performed by dividing one frame and processing each part).

On the other hand, in the other case where beamforming is performed a plurality of times at different times, the timing of transmission from the first excited aperture element in the effective aperture array can be grasped in the channel of each reception aperture element. In the device according to the embodiment, in the digital signal processing unit, a plurality of received signals can be appropriately overlapped to obtain the same signal as the received signal when a plurality of waves are transmitted at the same time, and processed in the same way ( Also in this case, parallel processing may be performed by dividing one frame and processing each portion). In these cases, substantially only one initial Fourier transform is required (there are cases where each part is divided and processed).

In such an active case, beamforming can be completed at a higher speed in beamforming other than aperture synthesis (methods (2) and (3)). However, the received signals for aperture synthesis collected for methods (2) and (3) are subjected to beamforming for transmission (transmission delay or transmission apodization is applied as necessary) and superimposed. and may be treated in the same way. By the way, if the received signals are superimposed without applying a transmission delay, in this case, a received signal can be generated when a plane wave is transmitted without steering. Also in aperture plane synthesis processing, division parallel processing may be performed to increase the speed. Beams in multiple directions can be generated from the signals collected in, and when processed by the apparatus or method of the present invention, the same angle obtained by a single Fourier transform under the deflection angle in each direction Calculations are performed using the spectral data, and finally, without performing multiple inverse Fourier transforms of the multidimensional spectra obtained for each deflection angle, the multidimensional spectra are superimposed to yield a single inverse A final image signal can be generated at high speed by Fourier transform (in some cases, division processing is performed). However, whether it is aperture synthesis, fixed focusing of reception, or other beamforming, when performing passive processing, as described above, for the superimposed received signal (that is, one angular spectrum), For example, it is useful to operate in a single direction or in multiple different directions.

In addition, when a plurality of waves obtained by these processes can be superimposed, for example, if the propagation direction, frequency, or band are different, it is effective to separate the spectrum and process it. , and others can be separated in digital signal units using coding, MIMO, SIMO, MUSIC, independent signal separation (independent component analysis), principal component analysis, code, or parametric methods. Note that the superposition process may be effective in other cases (for example, using a plurality of signals obtained in the same phase improves the SN ratio, etc.).

Independent signal separation (independent component analysis) is useful, for example, to separate the specular and scattered signals, for those containing the same specular signal, between two or more frames, independent scattered signals. By realizing and processing a state with or mixed with independent scattering signals, the commonly existing specular reflection signals can be effectively separated. It is useful for automating the detection and separation (removal) of high-intensity signals from blood vessels, etc., and the identification (detection) of the range of blood flow regions when measuring tissue displacement such as blood flow. be. In addition, it is useful for detecting and extracting boundaries of organs, tumors, etc. Similarly, specular reflection (tissue) can be detected, separated (removed), and the range of characteristic regions can be specified (detected). . At the same time, it is also possible to separate the mixed independent scatter signals, rather than detecting their respective signals from each of the sums (i.e., averaging) or differences of signals between frames, independent component analysis (independent component analysis) Detectability of the specular reflection signal and separation ability of the mixed-in signal are higher when using the component separation. Those capabilities are higher if they are processed after being detected (envelope detection, square-law detection, absolute value detection, etc.). This can be confirmed not only visually, but also quantitatively evaluated in a definitive or probabilistic manner. In addition, by applying the measurement of displacement, strain, etc., performing motion compensation (translation, rotation, expansion and contraction) with respect to translation, rotation, deformation, etc., and performing processing by performing alignment, these capabilities are improved (for example, 3 MHz In the frequency simulation, when the cross-correlation-based displacement measurement is performed, the specular reflectance distribution of 0.1 to 0.5 when the standard deviation of the scattering signal is 1.0 can be obtained even if the scattering signal of the same intensity is mixed. motion compensation). Processing with spatial resolution is desirable. It should be noted that the measurement of the displacement or the like is more accurate if it is performed before the wave detection, but it may be performed after the wave detection. Also, when performing highly accurate displacement measurement using various measurement methods before detection, block matching (phase matching) in the spatio-temporal domain that may be performed through oversampling or upsampling, or in the frequency domain Motion compensation by phase matching based on phase rotation is effective. An independent signal can be obtained by tilting the transducer to pick up the specular signal at the same position from another angle, or picking up the signal using a different sub-aperture based on the steering process, if it is a medical ultrasound transducer, and It is also effective to shift the scan plane and collect signals including specular reflection signals from the same specular reflection signal source (sequence of the same structure and composition) (measurement technique). It is also possible to actively apply the movement of the object (scanning plane shifts or signals from other tissues are mixed) and deformation (considered to be a state containing noise), and as described above, the specular reflection signal It is sufficient to collect signals containing Similarly, when using ultrasonic waves (such as sonar) or electromagnetic waves other than medical ultrasonic waves, movements of sensors, signal sources, detectors (including camera shake and disturbance of their holders), waves or Beam steering, object motion and deformation may be applied to collect and process reflected or transmitted waves. Noise generated in the circuit may be mixed in to bring about effects, or intentionally generated analog or digital noise (including software noise by a program) may be mixed in. These processes can be used not only to separate the specular reflection signal and the scattering signal, but also to obtain the same effect on the common signal and the mixed signal between the signals, and the range of application is limited to this. not a thing The spatio-temporal deviation of the signal is not necessarily due to displacement or distortion, but also the directivity of the aperture (especially during steering), the inhomogeneous propagation speed of the medium itself, the disturbance of the medium, and changes in conditions (e.g., pressure and temperature changes). etc.) may also be processed purely in signal analysis. In addition, waveform distortion may occur due to frequency-dependent phase delays and phase changes in devices (for example, phase changes in sensors and phase delays in amplifiers in circuits). The correction may shape the waveform (it is useful to rotate the phase of each signal component by multiplying it by a complex exponential function frequency by frequency in the frequency domain). Obtaining the phase change occurring in the device in advance to generate a wave having a desired waveform, or appropriately electrically driving the sensor with the goal of generating a desired waveform at the observation position or receiving position. There is also As a result, the observation of the observation target can be made with high accuracy (various observations using phase, observations of frequency-dependent propagation velocity, scattering, attenuation, etc.). Only the amount that occurs at the receiving device may be corrected. Other misregistrations mentioned above may also be processed in the same way. Although the frame signal has been described, it may be applied to beamforming (including aperture synthesis), or before reception beamforming or without beamforming at all (transmission and reception signal for aperture synthesis) Beamforming may be performed after processing the That is, it may occur before, during, and/or after beamforming. In each case, super-resolution may be used in combination. In addition to correcting spatio-temporal shifts, the motion compensation process described above is also effective in correcting shifts when comparing and referring to signals when transmitting focusing beams or plane waves (for example, in super-resolution). . Further, the above motion compensation processing that may be performed before or during beamforming may also serve as delay processing in DAS processing. In addition, detection (absolute value detection, square-law detection, envelope detection, etc.) and high resolution through linear or nonlinear processing, which will be described in detail later, are also applied to the beamformed (including aperture synthesis). In some cases, beamforming is performed before receiving beamforming or after performing processing on a signal (transmit/receive signal for aperture plane synthesis) for which beamforming has not been performed at all. That is, it may occur before, during, and/or after beamforming. When band widening is performed in these processes, oversampling or upsampling is performed in the space and time, or band widening is performed by zero-filling the spectrum in the frequency domain (inverse Fourier transform is performed) as necessary. For example, oversampling or upsampling results are obtained).

For the position correction and phase aberration correction described above, a cross-correlation method can be used as a correlation-based method, and various displacement measurement methods can be applied. It may be processed as a one-dimensional signal, or it may be processed as a multi-dimensional signal. For example, there are (multidimensional) cross-spectral phase gradient method, (multidimensional) autocorrelation method, (multidimensional) Doppler method, and the like. Similar to the cross-correlation method such as the cross-spectrum phase gradient method and the autocorrelation method, if the matched filter has an effect, it becomes robust against noise contained in the wave signal as described above. As mentioned above, it is advantageous to use (iterative) phase matching methods (shifts in space-time or phase rotations in frequency space) when applying these displacement measurement methods to phase aberration measurements. Here, an example of a method for automatically determining phase aberration and effective aperture width in beam forming using an array type aperture applying these will be shown. As shown below, discrimination based on the correlation value between local signals (for example, local echo signals and local transmitted ultrasound signals in the case of ultrasound) and detection of occurrence of delay time (time difference) estimation error itself in beamforming By doing so, it is possible to automatically determine the effective aperture width and perform beam forming. In that case, the iterative phase matching method is effective. For example, as described above, if it is applied to coherent compounding in beamforming with different steering angles, imaging can be obtained, and effects such as increased contrast and higher resolution can be obtained. In addition, as described above, MIMO, SIMO, MUSIC, independent signal separation (independent component analysis), principal component analysis, parametric methods, etc., adaptive beamforming, minimum variance beamforming, Capon method, etc. (non-patent literature 10) can also be applied. Not only are these highly accurate, but the amount of calculation can be reduced. The adaptability itself depends on the reflection characteristics, scattering characteristics, or attenuation characteristics of the measurement object, and it is desirable to have spatial resolution.

At least received signals that have not been subjected to receive beamforming will be described (as described above, signals subjected to receive beamforming may also be processed). Examples of transmission beamforming include fixed-focus beamforming, plane waves, spherical waves, and other waves that spread widely in a direction intersecting the propagation direction of the waves (but not limited to these). The latter achieves a high frame rate as described above. Moreover, when transmission beamforming is not performed, this method is used for so-called classical aperture synthesis, and may be effective in both monostatic and multistatic types.

In those signals, in the signal sequence received at each position around the reception position (position where beamforming is performed) that received the local signal including the signal from the position of interest in the region of interest Find the highest correlated local signal present and its time of reception (see FIGS. 16-18). A search area for that purpose is set so as to include highly correlated local signals. High-precision phase rotation by multiplying the complex exponential function in the frequency domain is applied to the signal in the search area, and the cyclically appearing signal is cut out with a window to obtain the phase aberration. The search area should be appropriately set larger than the local area so that it does not appear in . appropriately determined from the magnitude and sign of the estimated phase aberration). If the phase aberration between the local signals is estimated to be Δt, exp{iωΔt} is added to the spectrum A(ω) of the signal in the search area to correct the phase aberration by shifting the signal in the search area by −Δt. multiplication and inverse Fourier transform (Patent Document 6, Non-Patent Document 15, etc.). Although computation time is required, if this process is repeatedly applied to the same pair of signals, the correlation between the local signals gradually increases, and finally a highly accurate estimation of the phase aberration can be achieved ( The iterative phase matching is detailed in Patent Document 6, Non-Patent Document 15, etc.).
FIG. 16 is a schematic diagram for explaining an example of phase aberration correction when a linear one-dimensional array transducer is used and steering is not performed. FIG. 17 is a schematic diagram for explaining an example of phase aberration correction when steering is performed when a linear one-dimensional array transducer is used. 16(a) and 16(b) show received signal groups obtained using different parameters, and FIGS. 17(a) and 17(b) show received signals obtained using different parameters. 2 shows a group of received signals. As shown in FIGS. 16(a) and 17(a), when processing is performed only within a frame consisting of a group of received signals including a point of interest A for which beamforming is performed, FIGS. As shown in b), received signal groups (frames) with different wave parameters and beamforming parameters are obtained, and the received signals at the point of interest A shown in FIGS. phase aberration correction. The figures illustrate the processing of multiple frames obtained at different transmit or receive steering angles.
FIG. 18 is a schematic diagram for explaining an example of phase aberration correction when steering is performed when a two-dimensional array type transducer is used. FIGS. 18(a) and 18(b) show received signal groups obtained using different parameters. This transducer is of the linear type, but there are various types of two-dimensional array type transducers as well as the one-dimensional array type, and it is basically used for transducers with arcs and for sector scanning. Some transducers (even non-linear transducers may steer about an axial direction determined by the direction in which the element aperture faces). Since the digital signal stored in the memory after reception is to be processed, if the cross-correlation processing itself is applied, the reception time of the highly correlated local signal will be evaluated at the sampling interval. The inventor has reported a cross-spectrum phase gradient method capable of analog evaluation of digital signals based on the Nyquist theorem, and this method can be used, for example. As described above, other displacement measurement methods can be applied to correct the phase aberration. In the case of iterative phase matching, the window is often a rectangular window, and it is especially desirable that a rectangular window is used at the end of the iteration), and there is no need to perform local estimation such as moving average, as shown in Fig. Estimation using the received signal (instantaneous data) of the interest point A shown in is also possible. In addition, the received signal to be processed referred to in this process may refer to a signal that has already undergone beamforming for transmission or reception.
This cross-spectrum phase-gradient method requires unwrapping because the phase spectrum is inverted when the time difference between the local signals is large. Therefore, so-called coarse estimation and phase matching (spatial shift) using the cross-correlation method itself including the unwrapping effect are first performed, and then fine estimation using the cross-spectrum phase gradient method is performed. In this fine estimation, phase rotation using a complex exponential function is performed as phase matching, and it is possible to improve accuracy while repeatedly improving the correlation value. can). This approach is based on the phase matching method established by us as a displacement measurement method (Patent Document 6, Non-Patent Document 15, etc.). The calculation speed of the cross-spectrum phase gradient method is faster than that of the cross-correlation method, so the cross-spectrum phase gradient method with unwrapping is also effective It suffices to perform unwrapping determined by the difference, which is easier than when observing displacement when moving in an arbitrary direction). In this iterative phase matching, it is also possible to gradually shorten the window length and finally obtain a high-resolution result. It is possible to coarsely search the range where the corresponding signal can exist, which is physically determined by the assumed propagation speed, and then perform fine estimation. In some cases, unwrapping is not required.
The effective aperture width is the direction away from the interest point A of each element position for beamforming in the aperture array (FIGS. 16 and 17: left and right in the case of a one-dimensional array, In the case of dimension, the above processing is performed on the received signal received at the position of (in the peripheral direction), and after the matching processing (i) before the correlation value obtained by the inner product between the local signals falls below the set threshold and (ii) before the time difference between the local signal and the obtained local signal becomes larger than a preset threshold (the estimated value increases discontinuously during the process). may change and may be considered to detect errors).
It is also effective to perform this processing when performing beamforming only with the received signal within its own frame. Since the wave motion has a narrow band in the horizontal direction, it is effective to perform this processing when performing coherent compounding processing (broadband) by performing a plurality of steering transmissions with different steering angles as described above. The effect of widening the band is also obtained in focusing beam forming, which is of course effective. Received signals obtained by performing beamforming with different beamforming parameters and wave parameters other than the steering angle are similarly processed (see FIGS. 16 and 17). The above processing can be performed and compounded at each position A of the received signal frame under a reception steering angle that is the same as or different from the transmission steering angle. may not be imaged. Therefore, after determining the transmission steering angle of the wave to be generated, one reception steering angle that is the same as or different from it is determined, and for each position A that is set only within the reception signal frame of that transmission steering angle, the own reception The same processing is performed not only on the signal frame (only this frame is processed when compounding is not performed), but also on the received signal frames of other transmission steering angles. When pursuing accuracy, the steering direction of transmission and reception should be determined in the direction of the aperture with strong directivity (that is, the front direction of the aperture). Furthermore, it is possible to generate and compound with different combinations of steering angles for transmission and reception, but if the steering angles for transmission and reception are greatly different, images may not be formed.
Regarding compounding, for example, for one transmission steering angle in the front direction or an arbitrary direction, multiple signals are generated at multiple different reception steering angles, or multiple directions are generated using synthetic aperture plane data. We also report steering and compounding of transmit and receive, which can be applied to transmit steering in multiple different directions as described above.
In this process, the length of the local signal (window length) must be appropriately set (this is the initial window length when performing iterative matching). The shorter the distance, the higher the correlation value obtained, and the higher the spatial resolution of the local signal estimation result (that is, the phase aberration estimation value), but there are cases where another similar signal is detected. If the window length is long, the correlation value becomes low, and the signal cannot be found (for example, when ultrasonic plane waves with a nominal frequency of 7.5 MHz are transmitted through a graphite agar phantom of human soft tissue, echo signals are transmitted at a frequency of 30 MHz). 64 and 128 points were appropriate, but 32 and 256 points were inappropriate). It is similar to observing displacements and displacement vectors. As described above, it is preferable to perform phase matching with an appropriate window length, and when performing iterative phase matching, it is effective to gradually shorten the window length while repeatedly performing phase matching. Yes (ultimately high-resolution results are obtained). Iterative phase matching is effective in the case of aperture plane synthesis and the like, in which scattered signals that are not imaged by phase matching several times are imaged by repeating the number of times. The highly accurate phase aberration correction or iterative processing described above requires computation time, but is suitable for precision inspection. Although medical ultrasound emphasizes real-time, the above processing has the potential to be a novel medical ultrasound work-up method. The iterative phase matching can be performed with an upper limit value, and is terminated when the updated value of the estimated value of the phase aberration becomes smaller than a preset value or becomes sufficiently small, and the above ( While the condition i) or (ii) is satisfied, the same processing is performed on the received signal at the next position in the horizontal direction. It is also effective to use the obtained estimation result for the initial value of the phase aberration at the next position in order to shorten the calculation time. In the case of a one-dimensional array type, it is possible to carry out the processing left and right with the point A as the center (the effective aperture width obtained by the above processing (i) or (ii) is obtained by not necessarily symmetrical about the center). In the case of the secondary array type, it is possible to perform the processing in the circumferential direction with the point A as the center. Alternatively, the maximum effective aperture width including the point A is determined in advance, the phase aberration is estimated from the end, and the correlation value of condition (i) is larger than the predetermined value and the phase aberration of (ii) is determined. It is also possible to determine the effective aperture width based on the aberration estimate itself. In any case, each time the estimation result of the phase aberration is obtained, the phase aberration-corrected local signal (delayed local signal) is obtained. It is efficient to perform addition processing (that is, DAS processing using phase aberration). In addition, at that time, independent signal separation (independent component analysis) and principal component analysis are performed, and signals with different components from signals with the same components as the reference signal (signals with higher accuracy than the above sum (additional average)) (a signal that is more accurate than the above subtraction) and performing an addition process on the former signal component. To discriminate between the former and the latter signal components, the correlation with the reference signal is calculated, and the signal with the higher correlation value is regarded as the former, and the signal with the lower correlation value is regarded as the latter. The latter signal component may also be used after being subjected to addition processing. Then, phase aberration correction can be similarly performed for another position A in the depth direction and the lateral direction within the region of interest. If the purpose is only to estimate the phase aberration, the addition process is unnecessary. In addition, independent signal separation is performed once on the signal group subjected to phase aberration correction within the effective aperture width estimated at each point of interest A, and a signal component common to all (signal with the highest correlation value) and other signal components, and the former may be used as the result of DAS processing.
In addition, when signals with different wave parameters or beamforming parameters for transmission or reception overlap, phase aberration estimation and phase aberration correction may be applied to the signals as they are. Alternatively, they may be separated by other processing, and then phase aberration estimation and phase aberration correction may be performed on each separated signal (for example, processing may be performed on each steering angle signal). . Similarly, the spectrum of overlapping signals and non-overlapping signals may be divided in the frequency domain to generate new waves, and then phase aberration estimation and phase aberration correction may be performed. When the signal is separated in the frequency domain, wave parameters and beamforming parameters can be realized, for example, pseudo-steered waves in various directions, new waves and beam shapes, etc. can be generated (but not limited to this) ). Signals with phase aberration correction in each may be superimposed.
It has been confirmed that the image contrast and spatial resolution are higher in the processing of individual frames than in the image without phase aberration correction, and multiple frames are processed and compounded as described above. has been confirmed to be even more effective. This compounding performs coherent processing on the raw signal, but after performing detection (envelope detection, square law detection, absolute value detection, etc.), superimposes (incoherent compounding) to reduce speckle It is also effective in generating an image with improved contrast and CNR (Contrast-to-noise ratio) by emphasizing a decisive signal such as a specular reflection signal or a strongly scattered signal.
If other methods such as the (multi-dimensional) autocorrelation method or the (multi-dimensional) Doppler method are used, the cross-correlation method can be used for the coarse estimation and can be used for the fine estimation without phase unwrapping. It may be used alone like the cross-spectral phase gradient method, or it may be unwrapped. The effective aperture width and phase aberration estimated as described above are super-resolution by minimum variance beamforming, independent component analysis, principal component analysis, and nonlinear processing (nonlinear processing is performed and addition processing is performed. It can be used for various applications including the above applications such as applying non-linear processing after applying. It is also possible to create a database of the average effective aperture width at each distance position based on method (i) or (ii) and use it without phase aberration correction. The above process is the same for the passive second embodiment.

The signal separation is performed by increasing the frequency and widening the band (when the order is greater than 1) or by reducing the frequency and narrowing the band (when the order is less than 1) by exponentiation, and then the frequency domain can be performed with high accuracy. Restoration of the signal after separation is also easy by multiplying the reciprocal of the exponentiation order used.

The phase aberration correction described above is also effective in various other cases. Phase aberration correction is possible even for various observation data (between), such as when the observed quantity (physical quantity, chemical quantity, etc.) is different, or when the conditions and parameters that affect the same observed quantity are different. may be applied, or phase aberration correction may be applied prior to various processing described herein. For example, when fusing and integrating ultrasound (echo) and OCT and observation data obtained by using them, it is possible to obtain characteristic positional accuracy from OCT data and correct the phase aberration of ultrasound (echo). (Although the physical properties of light and sound are different, the position where they change and the position where scattering occurs can be used in common, and the matching process itself may be performed between coherent signals, or detection may be performed. It is sometimes performed between incoherent signals or between image data obtained by the method). After matching, for coherent signals or incoherent signals (or image data), as described later, in addition to ICA, machine learning, deep learning, neural networks (back propagation type, Hopfield type, etc., or Other models such as modified models based on these) or other processing may be performed.
The present inventor is mainly involved in medical imaging and remote sensing (various types of radars, terrestrial radars, extraterrestrial radars, stars and satellites, airborne, meteorological observations, various earth observations, environmental and resource exploration, space and astronomical observations). , sonar, etc.), non-destructive inspection (structural science, material research, physical properties and functions, including life), and other breakthrough multidimensional digital signal processing technology (electronics Based on information processing including inverse analysis/inverse problem, statistical processing (based on mathematics), we are working on the fusion of measurement technology and advanced information processing. In particular, we will create innovative science and technology that can be applied comprehensively in the fields of human tissue (including microscopy), bio, materials, structures, and the environment, and bring about social and economic transformation. becomes possible. For example, observation results can be obtained numerically, graphed, imaged, visualized, medical and life sciences (human health promotion, etc., human tissue examination, generation, treatment, processing, application, (small) animals, tissues, cells, drugs, etc.), development of materials (electric, thermal, elasticity, composite materials, etc.), environmental observation (gases, liquids, solids), energy development, security (surveillance and monitoring, observation of moving objects) , various high-precision observations (large and small objects, short-term and long-term phenomena, distant and short-range objects), and the world's first in silico measurement standard can be realized. Three basic physics of electromagnetism, mechanics, and heat that can be applied to a wide range of applications, and other phenomena (biology, chemistry, biochemistry, etc.) Innovative non-destructive inspection techniques for physical quantities, chemical quantities and physical properties in various phenomena) can be realized, and measurement (inspection and diagnosis), repair (repair, treatment, regeneration), manufacturing (materials growth and three-dimensional tissue culture) and applications (including the creation of new functions and physical properties). This will enable in situ measurement of basic physical quantities, chemical quantities, and spatio-temporal distributions of physical properties in objects that could not be observed so far, and it will be possible to observe without preparing observation conditions. For example, it will be possible to observe working devices, living organisms, and cells under natural conditions, and to observe the growth process of materials under conditions. In addition, various latent factors (mechanisms), new phenomena, new principles, etc. can be elucidated in detail through the advanced fusion of these diverse simultaneous observation and analysis technologies, cutting-edge information science and statistical mathematics, etc., and new physical properties and It can contribute to the creation, synthesis, and restoration of functions. Rather than simply over-determined systemization (additional averaging and least squares) to exceed the accuracy limit of single observations, for example, the following intelligent integration and fusion technology (multiplexing of high-precision common information and Separation of independent information, separation of independent signal sources) can dramatically exceed new applications of ICA, machine learning and deep learning, and new applications of neural networks such as integrated learning and integrated recognition). Of course, if necessary, the sensor can be dismantled, and in humans, craniotomy, laparotomy, laparoscope, endoscope, oral/nasal hole (camera), capsule (type camera), puncture needle can be used to observe the sensor. It may be observed in preparation for the vicinity. Also, phase aberration correction may not be performed at all, or may not be performed precisely (for example, when using the cross-correlation method, the shorter the sampling interval, the better). The signal source mentioned here may be the physical signal source itself, or may be a reflection source or a scattering source. In particular, when the received signal includes signals of multiple reflections or multiple scattering, etc., when it is thought that the signals to be observed are included at different positions and times, the signals at different positions (time) are actively used for ICA, etc. It is effective to apply the processing of , and such processing may also serve as signal detection processing, such as processing while shifting the position (correlation value can be used as an index of the accuracy and reliability of the detected signal) . As a result, signals of multiple reflection and multiple scattering can be separated.

For example, as described in paragraph 0380, in situ imaging by wave sensing (including inverse analysis) of the distribution of physical quantities in electromagnetics, mechanics, and thermology in the observation object is realized, and based on those observations, related Reconstruction (inverse analysis) of physical property distribution can be realized. The source of the phenomenon can also be reconstructed. Also, one physical quantity and its reconstruction can be used to obtain another physical quantity. It is also possible to obtain and observe energy. In some cases, three-dimensional or two-dimensional multidimensional observations are desirable in order to achieve high-precision observations. for example,
(1) electrical properties based on current density (vector) distribution measurement (both conductivity and dielectric constant, or either one) distribution, current source or voltage source distribution, electric potential distribution reconstruction (imaging), Or electrical properties based on potential distribution measurement, current density (vector), current source reconstruction, etc. (2) Mechanical properties based on distribution measurement of displacement (velocity, acceleration) vector and strain (rate) tensor (shear modulus Viscosity modulus, compressibility (Poisson's ratio) and incompressibility, viscosity, density, etc.) distribution, reconstruction of force source distribution (imaging), mean vertical stress (internal pressure) distribution are simultaneously or separately observed, inertial force vector and Observation of stress tensor distribution, etc. (3) Reconstruction of acoustic properties (bulk modulus, viscoelasticity, density, static pressure, specific heat, etc.) distribution, sound source distribution, sound pressure, particle displacement/velocity distribution based on sound wave propagation measurement (Imaging), etc. (4) Thermophysical properties (thermal conductivity and heat capacity, thermal diffusivity, perfusion, convection) distribution, heat source distribution, and heat flux distribution reconstruction (imaging) based on temperature distribution measurement. Those observations in the observation target that is the medium of each wave are performed by nuclear magnetic resonance imaging (MRI: Magnetic Resonance Imaging), SQUID (Superconducting Quantum Interference Device) (observation wave is magnetic field), terahertz (electric field), direct current ( waves), electromagnetic waves such as electric power, radio waves, microwaves, infrared rays, visible light, ultraviolet rays, radiation, cosmic rays, sound pressure distribution observation by Doppler effect using optical pulses and lasers, OCT (Optical Coherent Tomography) , an optical microphone (inventor: Mr. Yoshihito Sonoda), an ultrasonic echo (sound wave) device (modality), and others (many patents of the inventor). The reconstruction of the physical property distribution of (1) to (4) is to reconstruct the relative physical property distribution from the distribution of one physical quantity, and many inverse analyzes (X-ray CT, electrical impedance CT, etc.) are observed. In order to observe the inside of the object from the physical quantity observed at the boundary position of the object, the integral equation is solved, whereas the partial differential equation or ordinary differential equation (the one in which the constitutive equation is substituted into the governing equation such as the wave equation or the diffusion equation) Since it is to be solved, it is called differential type inverse problem / inverse analysis as opposed to integral type. An absolute distribution can be obtained by giving a physical property reference value (measured value, typical value, etc.) to an appropriate position (reference region) within the region of interest. If a region with a constant value is used as a reference region and a unit size value is set as the reference value, a relative value distribution can be obtained. Of course, their integral inverse problems may also be used.
In any case, the equation Ax=b for the unknown distribution (vector x) holds (the number of unknowns and equations may be equal, over-determined, or less-determined). In the case of linear equations, the unknown distribution x is often the distribution that we want to find. It is often the distribution Δx. Of course, it is not limited to that. In the case of nonlinear problems, the matrix A or vector b depends on the unknown distribution x, and A(x) or b(x ) is used to find Δx at that step. Then, x is updated by Δx. When the magnitude (norm) of Δx becomes smaller than a predetermined value, it is determined that convergence has occurred, and the repeated estimation is terminated.

Furthermore, it is possible to fuse and integrate those electromagnetic wave and sound wave sensing and inverse problems. In the field of medical imaging, devices combining X-ray CT (morphological information), MRI (morphological and functional information), and PET (functional information) have been put to practical use. Under such circumstances, in addition to ultrasonic waves (longitudinal waves) and shear phenomena (transverse waves), we will make use of the characteristics of various waves, such as MRI and ultrasonic waves, ultrasonic waves and terahertz, OCT, laser, etc. It can also fuse with electromagnetic waves, mechanical waves, and heat waves.
For example, there are various applications of fusion and integration of (1) to (4) (hereinafter referred to as (5)) as described in paragraph 0380. For example, in the field of life sciences, new measurements of functions and physical properties can contribute to highly efficient culture (especially three-dimensional culture), its control, and elucidation of the mechanism of lesion development (simultaneous observation of organic and inorganic substances, etc.). ). While iPs cells are attracting attention, for example, it is possible to simultaneously observe the activity of both kinetic and electrical phenomena in the in situ state where cultured cardiomyocytes are naturally active (multiplexing and fusion of new observations are also possible). In addition, electrical and electronic circuits in operation can be observed in situ (device function and junction inspection). In addition, in the process of growth and operation of various functional materials and in the development of various devices, physical quantities and physical properties are observed in situ (for example, ultrasonic elements such as piezoelectric PZT elements and polymer film PVDF). At the same time, the energy conversion efficiency, electrical impedance, vibration mode, and thermal phenomena of electrical machines can be observed. materials, elastic bodies, thermal materials, etc.). In addition, effective observations are made in various fields as described in paragraph 0380.
Information science and statistical mathematics are effective in their reverse analysis and fusion. As described in paragraph 0381, various optimizations that have been performed in signal processing, maximum likelihood estimation, Bayesian estimation, EM (Expectation-Maximization), unbias regularization with a partial differential operator as a regularization term (in the past Implementing the world's first spatio-temporal variant regularization, setting a large regularization parameter in absolute or relative physical property reconstruction for homogeneous materials to stabilize as much as possible and reduce variability It is possible to realize the world's first in silico standard), singular value decomposition, equalizer and sparse modeling, signal (source) separation and feature value analysis by ICA and MUSIC, new super resolution (in silico harmonic imaging, etc.) patent application), and in the evaluation of the displacement (vector) measurement error, we have assumed a stationary process and used the Cramer-Rao Lower Bound (CRLB) and Ziv-Zakai Lower Bound (ZZLB) (a priori and a Also applied to posteriori regularization). It is also possible to establish an error model for each observable and apply it to them. In addition, for the integration and fusion of heterogeneous information (including the above-mentioned different observables, observables under different conditions and parameters, etc.), KL information, maximum likelihood method, mutual information minimization/maximization, and entropy minimization are used. In addition, in order to improve the multiplexing of common information and the ability to separate independent information, new ICA applications (precise spatio-temporal matching of multiple digital signals (digital If the signal is processed after applying analog phase rotation, etc. without approximation processing, the ability to extract common information exceeds that of arithmetic averaging, the ability to separate independent components is improved), and new deep techniques using neural networks Integrated learning and integrated judgment (recognition, approach to integrate and analyze multiple three-layer neural networks, differential diagnosis and recognition of lesions, etc. (innovative encoding of lesion types, encoding of various recognition targets not limited to medical diagnosis possible), mapping between images or other clinical data, desired objectives/goals, new functions, etc. (also not limited to medical applications). Integration and fusion of phenomena with different stochastic processes are also effective (for example, multiplexing or separating noise following a normal distribution and ultrasonic scattering signals following a Rayleigh distribution, mixing a plurality of different stochastic processes, etc.). Terahertz signals and SQUID signals have a low S/N ratio, and similarly, it is effective to improve accuracy by signal processing using ICA (exceeding averaging) and MUSIC. In addition, when sensing data in three-dimensional space or space-time is big data, in pursuit of real-time performance (high-speed calculation), precise and high-speed Fourier beamforming (including data compression) and parallel processing are required during sensing. Processing may also be performed. Of course, so-called data mining is also effective. Of course, various conventional compression techniques are also effective.

In the future, new observation techniques that are developed in parallel or integrated and merged will establish a position as an innovative non-destructive inspection technique for each observation target, and will open new prospects (new engineering) in various fields. etc.), and is considered to be a breakthrough in the development of science. In addition, it can also contribute to international measurement standards for basic physical properties (especially distribution constants) (it has the potential to become the first in silico international standard that uses regularization, etc.), and is of great industrial importance. effect (achieving precision equivalent to the number of significant digits of a computer). The above-mentioned in situ remote sensing application (approach) is not limited to these, and the ripple effect to other fields is immeasurable, and as described above, it can be returned to various fields, fusion/integration fields, and society.

Examples of applications realized by the above (1) to (4) are itemized below. Applications are not limited to these. In addition, the distribution of changes in the thickness of the object to be observed can be observed with high accuracy from the reconstruction of the distribution of physical properties (thermal conductivity and electrical conductivity). At least one value may be observed without being a distribution.
(1) electrical properties based on current density (vector) distribution measurement (both conductivity and dielectric constant, or either one) distribution, current source or voltage source distribution, electric potential distribution reconstruction (imaging), Alternatively, electrical properties, current density (vector), and current source distribution reconstruction based on potential distribution measurement.
Platform: MRI, terahertz, SQUID, electrode array (electroencephalogram, electrocardiogram, etc.), electrostatic potential sensor array, etc.
Observation target: MRI, electrode arrays, and electrostatic potential sensor arrays are human and animal (mouse, etc.) neural networks, electrical networks, electrical materials (resistors and dielectrics), current fields, electric fields, and potential fields; Terahertz is electric circuits, electric materials (resistors and dielectrics), piezoelectric elements (PZT, PVDF), etc.; SQUID is all of them. When MRI or SQUID is used, the current density vector distribution can be found by solving the inverse problem of the Biot-Sabbat law (integral inverse problem).
(2) Mechanical physical properties (shear modulus, viscous modulus, compressibility (Poisson's ratio, etc.) and incompressibility, viscosity, density, etc.) based on distribution measurements of displacement (velocity, acceleration) vectors and strain (rate) tensors Reconstruction of distribution and force source distribution (imaging), simultaneous or separate observation of average vertical stress (internal pressure) distribution, observation of inertial force vector and stress tensor distribution, etc.
Platform: MRI, terahertz, ultrasound, OCT, laser, optical pulse, etc.
Observation target: Human and animal (mouse, etc.) soft tissue (brain tumor, cancer lesions such as liver cancer and breast cancer, vascular disease models such as sclerosis, heart and blood vessels, hemodynamics including blood flow, especially MRI Intracranial cancer lesion models that are difficult to observe with OCT); Viscoelasticity, liquids (including blood) containing powders (tracers) such as drugs and metals (conductors and magnetic substances), fluids such as gases, gases including dust, waste liquids, etc.
(3) Reconstruction imaging of acoustic properties (bulk modulus, viscoelastic modulus, density, static pressure, specific heat, etc.) distribution, sound source distribution, sound pressure, particle displacement/velocity distribution, etc. based on sound wave propagation measurement.
Platform: optical pulse, optical microphone, laser, MRI, ultrasound, OCT, etc.
Observation targets: In addition to the above organic observation targets, gases (helium, oxygen, air, etc.), liquids (pure liquids, mixed liquids, water, saline solutions, liquids containing drugs, blood, etc.), solids (inorganic systems, etc.) include). Special environments (indoors, outdoors, high altitudes, mountains, depths, oceans, narrow spaces, etc.), etc.
(4) Thermophysical property (thermal conductivity, heat capacity, thermal diffusivity, perfusion, convection) distribution, heat source distribution, reconstruction imaging of heat flux distribution, etc. based on temperature distribution measurement.
Platform: MRI, terahertz, ultrasound, optical fiber, pyroelectric sensor, etc.
Observation target: Thermal materials, inorganic solid piezoelectric materials (terahertz), etc., perfusion effects in neural networks of animals (mice, etc.), metabolism, cancer lesions and inflammation mentioned above, heart and blood vessels, blood (perfusion) models, piezoelectric elements , PVDF, gas and liquid convection, etc.
In addition to (1) to (4), as (5), (1) to (4) and other reverse analyses are merged and integrated, or data mined. For example:
(i) Simultaneous or multilateral observation of the same object, or simultaneous or multilateral observation of multiple objects related to the same event (phenomenon) by (1) to (4) using the same or different equipment.
・In situ observation in medical care (health checkup, medical checkup, medical examination, diagnosis of disease, monitoring in various treatments, simultaneous or multilateral observation of the same organ, tissue, lesion, single disease or concurrent disease, or Multiple independent diseases, related organs and tissues can be observed simultaneously or multilaterally, and these can be implemented with a small number of types of equipment (hardware).
・In situ observation in regenerative medicine and life science (highly efficient culture, especially 3D culture and its control, elucidation of the mechanism of lesion development, simultaneous observation of organic and inorganic substances, etc.).
・It is possible to simultaneously observe both kinetic and electrical phenomena (activity, etc.) in the in situ state where cultured cardiomyocytes (iPs cells, etc.) are naturally active, and to multiplex, improve accuracy, and fuse new observations.
・(Visco)elasticity (including hemodynamics) and thermal properties (and temperature) of blood (including drugs) in human and mouse models of lesions (brain tumor, liver cancer, breast cancer), myocardium, blood vessels, heart chambers, and blood vessels Simultaneous observation of
・Ultrasound and terahertz observation of drug delivery by dynamics and electromagnetic induction, development of new drugs.
・Simultaneous observation of electrical activity, heat generation (temperature), and blood perfusion effects in neural networks and metabolism in humans and animals.
・In minimally invasive hyperthermia treatment (HIFU: Heating or heating using High Intensity Focus Ultrasound, etc.) for human cancer lesions and Parkinson's disease, therapeutic effects (temperature rise, viscoelasticity change, etc.) of the affected area by MRI and ultrasound Simultaneous observation of denaturation), perfusion itself, and neural network control of perfusion, real-time multifaceted monitoring of treatment to improve therapeutic performance And to efficiently treat (repair) only the abnormal site.
・Integrated medicine (comprehensive diagnosis, treatment, surgery, physical/chemotherapy, medication, etc.), theranosis We have succeeded in consistently monitoring under the same index before, during, and after diagnosis and heat treatment, which is considered a successful example of theranosis), and changes in viscoelastic modulus due to inflammation after treatment (blood Simultaneous observation of current).
・Confirm the Wiedemann-Franz law for conductivity and thermal conductivity, and provide high-precision or low-cost techniques (for example, use an infrared camera instead of a SQUID meter).
・Optimization processing of materials, structures, production processes, etc., targeting desired characteristics in synthesis problems (composite materials, etc.).
・Contribute to the creation and synthesis of new functions by observing the energy conversion efficiency, electrical impedance, vibration mode, and thermal phenomena of electric machines at the same time in ultrasonic devices such as piezoelectric PZT devices and polymer film PVDF. (Electrical materials, elastic bodies such as rubber, thermal materials, etc.).
・Spontaneous movement of the robot by in silico integrated judgment.
・Relatively inexpensive (for example, a small number of devices, such as one, can be used for many purposes).
・For example, in the medical field, the distribution of physical properties such as (visco) shear elastic modulus is observed by observing the distribution of displacement (vector) and deformation of the observed tissue using ultrasonic echo method, MRI, OCT, laser, or light pulse are reconstructed, and the original images and their observation results are fused to diagnose and differentiate lesions (in addition to various fusion and integration processes, in images, multiple observation data are simply superimposed by watermarking at the same time). When performing some kind of treatment (surgery or physical/chemotherapy), even the therapeutic effect (mainly degeneration) is diagnosed using the same index (observed physical properties), i.e. , It is effective to diagnose, monitor, and observe the progress before, during, and after treatment. Using the chemical shift of the nuclear magnetic resonance frequency (Larmor's frequency), using the temperature dependence of light (refractive index, etc.) when using light such as OCT, and the temperature of the observed (visco) shear elastic modulus It is possible to observe the temperature (change) distribution using dependence, etc., and monitor the treatment effect. Estimation of the heat source (by obtaining the heat source based on the inverse problem, or by observing the heating / warming wave by sensing and obtaining the autocorrelation function, the shape of the heat source can be grasped, and the power of the heat source can be calculated in consideration of the transmission power and tissue physical properties It is possible to estimate or utilize the shape data for the inverse problem, and the distribution of the wavelength and propagation velocity of the wave can also be observed using the above autocorrelation function) (Patent Document 11), which is generated by heating It is possible to estimate and predict the temperature distribution, make an integrated judgment while observing them, and sequentially create a heating and heating plan to realize minimum-invasive treatment. If the receiver is capable of receiving during HIFU transmission, it can process and use the echo signal generated by HIFU transmission, and if reception is not possible, it can obtain a received signal by transmitting ultrasonic waves for observation. There is also a thing. Both can be used for imaging. Phase aberration (caused by temperature dependence of sound speed, inhomogeneity of sound speed, wave directionality, etc.) is calculated using at least one of the received signals, and phase aberration correction is applied to HIFU treatment and observation imaging. may be applied to improve the positioning accuracy of the treatment position. Based on observations and predictions, iterative HIFU beam and treatment parameters may be determined by optimization (various linear and non-linear optimization methods can be used). Linear or non-linear programming methods are also valid. It may be optimized by targeting a desired temperature distribution, exposure dose, tissue physical properties such as (visco)elastic modulus, and tissue pressure. At that time, typical data, measured values, or models of the heat-receiving properties and degenerative properties of the tissue to be treated may be used. The present invention is based on the above observations and predictions realized by making full use of the signal processing described in the present application, or optimization, HIFU irradiation power, irradiation intensity, continuous irradiation time, irradiation interval, irradiation position (focus position), irradiation shape (apodization), or HIFU interval, etc., electronically or mechanically controlling HIFU treatment to achieve Minimum-invasive treatment. Treatment control itself may be manually controlled, with the clinician making decisions based on observations, predictions, or optimization results for clinician-marked lesions or mechanically diagnosed lesions. In some cases, it may be controlled automatically mechanically, but in the latter case, it is necessary to always allow the clinician to change the treatment strategy and switch to manual control mode. need to make it possible. Switching from manual mode to automatic mode may also be useful. In either case, lesion tracking is important (cross-correlation-based processes such as cross-correlation and cross-spectral phase-gradient have been shown to be robust to changes in ultrasound images caused by heating (noise sources). ). The interface may be realized mainly by a PC and peripheral devices, or may be realized as a dedicated machine.
This method is one of the means to realize theranosis, has extremely high spatial resolution for diagnosis as well as treatment, and minimizes invasiveness under the differentiation of diseased tissue, nerves, blood (vascular), lymph, and niche. Realize the best possible treatment. Application to microsurgery is also important. In addition to early accurate diagnosis, a single device can be used to simultaneously diagnose diseases of a single or multiple organs (brain, liver, kidney, breast, prostate, uterus, heart, eye, thyroid, blood vessel, skin, etc.). It enables integrated diagnostic imaging that can be performed, and early and precise HIFU treatment, even under the same index (mechanical and thermodynamic physical quantities, tissue viscoelasticity and thermophysical properties, or markers, etc.) Diagnostic imaging and monitoring of therapeutic effects, as well as follow-up after treatment, are possible. We can realize the medical care that can be implemented. Under the development of medical technology suitable for the super-aging society, it is a medical technology that will be useful for a long time in the future while opening up a new clinical style (short-time diagnosis and treatment, examination, etc.). These means are effective not only in HIFU but also in radiotherapy and heavy ion radiotherapy. Combinations thereof, including drugs, are also effective. Also, waves for observation are not limited to ultrasonic waves, and various sensing waves such as MRI, OCT, and X-rays can be used (many others are described). These are sometimes used in combination and integration. In addition to medical applications, similar observations may be effective when performing diagnosis, repair, creation, etc. in material engineering.
・Elucidation of the functions of human and animal brain tissue: In cultured neural networks, the processes of learning and cognition, the effects of drug administration, etc. can be studied by electrical in situ observation, blood vessels in the heart and brain (viscoelasticity), and blood flow ( Fluids), microflows, etc. can be simultaneously observed with high precision, and in conjunction with these, external stimulus tools and processing techniques for those tissues and cells can be developed.
(ii) Application to the environment, industrial biotechnology, energy conservation, and environmental conservation (recycling, observation of air, soil, water, etc.). Acoustic physical properties of various gases (optical pulse, optical microphone), gases containing various dust particles (flow observation by terahertz), etc.
(iii) Development of new synthetic theories such as equivalent media and functional substitution (Materials Engineering).

Next, the above (1) to (4), other reverse analysis, and the fusion/integration means of (5) will be itemized. Techniques related to information science and statistical mathematics, including inverse analysis. Basically, it is intended to exceed the normal measurement limits. For example, the following processing is valid.
(A) Inverse analysis: In wave sensing,
・Using equalizers and sparse modeling of systems including observation targets (identification, dimensional reduction, rough sampling data conversion based on downsampling (regularization effect in inverse analysis, band compression in Fourier beamforming, etc.) It was faster, stabilized, and data compressed.
・ When performing inverse in measurement / reconstruction by optimization (weighted least squares, Bayesian estimation, maximum likelihood estimation (with or without MAP), singular value decomposition method, linear / nonlinear programming method, convex projection, etc.)・Integral inverse problem such as obtaining current density (vector distribution) data and current (distribution) from magnetic field (vector distribution) data in inverse analysis of Sabert's law) and physical property reconstruction (linear or nonlinear differential inverse problem) Problem: Stabilization of solutions by optimizing reference regions (initial values) and integral inverse problems (nonlinear inverse problems similar to impedance CT, etc.) in initial value problems using partial differential equations.
・Unique existence of solutions and adaptive stabilization by unbiased regularization when performing inverse analysis , there is a possibility that the world's first in silico standard can be realized by setting a large regularization parameter to stabilize the observation target as much as possible and reduce the variation. (implemented by modulus reconstruction, etc.).
・In addition to each observation target (strain tensor, temperature, current density vector, physical properties, etc.), an error model (variation and variance) of directly observed sensing signals is established to improve accuracy (in the past displacement (vector) In the evaluation of measurement errors, we assume a stationary process locally spatio-temporally, use Cramer-Rao Lower Bound (CRLB) and Ziv-Zakai Lower Bound (ZZLB), specifically when regularizing the system For a priori and a posteriori, regularization parameters are used so that they are proportional to the variance (thus, they can be spatiotemporally variable). The reciprocal of the variance and the reciprocal of the variance are used for each equation to weight the reliability (weighted least squares), and regularization and weighting can be performed at the same time.
・Super-resolution (higher resolution, the world's first in silico harmonic generation imaging, instantaneous phase imaging, well-known inverse synthetic aperture and inverse filtering), new minimum variance beamforming (phase aberration correction is effective), the above signal (source ), etc. Maximum likelihood estimation and the like have long been used in image processing, and for example, point spread functions that can be estimated in various ways, such as obtaining an autocorrelation function, are used (eg, Non-Patent Documents 31 to 34, etc.). There are various other methods.
・Analog or digital linear or non-linear processing of wave signals, generation and utilization of new waves by linear or non-linear phenomena in the propagation process (inside or outside the observation object), multiple waves with single or different parameters, This includes cases where objects with different observable quantities (physical quantities or chemical quantities) are the targets.
・Signal (source) separation and feature value analysis (independent component analysis ICA, principal component analysis PCA, MUSIC, application of the above regularization, singular value decomposition, machine learning, neural network, deep learning, etc.)
It is not limited to these.
again,
(B) Integration and fusion of different types of information, multiplexing of the same information and separation of independent information High-precision multiplexing of common information and separation of independent information: In signal processing and image processing, especially MRI (electromagnetic waves) and ultrasound (mechanical waves), ultrasound and terahertz (simultaneous observation of organic and inorganic), ultrasound In addition to using KL information, maximum likelihood estimation, mutual information minimization/maximization, entropy minimization, etc. for multiplexing and separation by new integration and fusion of sound waves (longitudinal waves) and shear waves (transverse waves), etc. In order to dramatically improve accuracy,
・ ICA (Past ultrasound echo signals (random signals) of human tissues were subjected to precise spatio-temporal matching of multiple data (phase rotation without approximation processing in digital signals) and then processed, common information The extraction ability of the
・Neural network (deep integrated learning and integrated judgment/recognition): An approach for integrated learning/integrated recognition of multiple three-layer neural networks with heterogeneous feature vectors (information) as input (multiple Connect neural networks, do not deep learn each neural network with closed data of each feature vector (information), let them learn to some extent and then combine them, and make them learn using closed data of all feature vectors (information) , the learning speed is dramatically faster than learning using closed data of all feature vectors (information) using a connection network from the beginning, and the recognition rate is also dramatically improved). The recognition rate is dramatically improved compared to the case where only a single feature vector (information) is used. For the integrated analysis, a novel single neural network that has already been trained and a joint network are used from the beginning of learning. Compare the weight distribution of the network when using it and when combining it after a certain amount of learning as described above and deep learning. It is effective to interpret independent components as . learning, observational data of the same or different types related to the recognition target, related observational data, or related recognition targets (e.g., diseases and lesions), desired purposes and goals, new physical properties and functions (created and synthesized by in silico and its utilization, and if it is difficult to make it into a device or if it is relatively simple and inexpensive, it may be operated in conjunction with other dedicated devices), etc.) (Learning and recognition of associative memory), other models that can be similarly implemented, optimization other than neural networks, and other processing described in this specification (in silico processing) can be similarly used.
・For phenomena with different stochastic processes (e.g., ultrasonic Rayleigh scattering, normal distribution of observation noise, etc.), multiplexing of common components, separation of independent components, and transition to different stochastic processes (stochastic models, changes in random variables, etc.).
again,
(C) Big data processing:
・High-speed and high-precision Fourier beamforming and parallel processing are implemented during sensing in order to achieve high-speed calculations in pursuit of real-time performance when handling big data such as large amounts of sensing data (spatio-temporal).
・Data mining (including the above-mentioned statistical processing, correlation processing, extraction of feature values by ICA, PCA, neural networks, etc.).
again,
(D) Improvement of observed signal SN ratio: For example, commonly observed terahertz signals, SQUID signals, light, etc. have low SN ratios, and can be processed by learning analog processing of correlation in OCT, digitizing and processing, ICA (exceeding averaging), MUSIC, Wiener filter, matched filter, correlation processing, signal processing such as signal detection (signal is real-time signal or complex signal).
Although the above processing and the like are performed in the present invention, the object and processing means of the present invention are not limited to these.

(1) to (4) of measurement/inverse analysis (imaging) performed using hardware (equipment, platform) and computers, and (5) fusion and integration (information science, statistical mathematics, etc.) using them , electromagnetism and mechanics, basic physics related to heat, other physics and chemistry, physical quantity and chemical quantity in biochemistry, non-destructive inspection of physical property distribution integrated and integrated in situ Innovative technology that realizes in real time (measurement and analysis method), and innovative measurement (testing and diagnosis), restoration (repair, treatment, regeneration), manufacturing (material growth and three-dimensional culture of tissue), application (new innovation by synthesis, etc.) in various fields creation of functions and physical properties, etc.). The present invention makes it possible to detect physical quantities, material states, their changes, latent factors, etc., which have not been captured so far, and to observe the situation in which the object to be measured is actually operating and functioning from various angles. , promotion of human health through medicine and life sciences, development of new engineering through the development of materials (including synthesis), food engineering (freshness and quality control, etc.), highly efficient energy development and resource exploration, energy conservation, environmental assessment , environmental conservation (including earth and celestial bodies), weather forecasting (weather forecasting, rainfall, gas convection, ocean currents, etc.), security, satellites, radar, sonar, international standards (potentially becoming the first in silico standard), etc. , has spread to various fields, and its ripple effects are enormous both economically and socially, including the industrial world and social infrastructure. The techniques developed will inevitably become basic technologies that contribute to the creation of scientific and technological innovation, the creation of new industries, and social contribution.

On the other hand, in the beamforming methods (1) to (6), the received signal stored in a memory or storage device (storage medium) in order to frequency-divide the spectrum and generate an image signal of one frame, the wavenumber Obtaining a plurality of processed waves in a situation where the spectrum is split in the frequency domain represented after matching may also be done. Angular spectrum states may be split and each processed. In either case, the band of the signal component should be limited for processing. Even when multiple waves overlap, the spectrum may be frequency-divided as well. It corresponds to the generation of physically pseudo waves with new wave parameters (frequency, band, direction of propagation, etc.) by frequency division of these spectra. Those split spectra may be processed in parallel. The superposition process is also a process to generate new wave parameters, but it may be superposed in the spatial domain and the angular spectra may be superimposed, or the spectra may be superimposed before the inverse Fourier transform. However, if necessary, the angular spectrum after Fourier transform or the signal after inverse Fourier transform may be superimposed. By applying the reversibility of the Fourier transform (Fourier transform and inverse Fourier transform), the generated signal is restored to the signal before reception beamforming (or before transmission/reception beamforming in the case of aperture synthesis), and other beamforming is performed. (e.g., different steering angles for transmission or reception, multiple waves with different steering angles for one transmission, etc.).
As super-resolution processing, spectral weighting processing (not so-called simple inverse filtering or deconvolution, but filtering to have a desired point spread function after performing these processing can be performed collectively) , non-linear processing, and instantaneous phase imaging excluding phase rotation are described, but may also be processed under DAS processing in addition to the present Fourier beamforming. For example, as described above, high-speed imaging is possible when transmitting waves that spread widely in the lateral direction, such as plane waves, circular waves, cylindrical waves, and spherical waves. It is effective to generate a plurality of waves with different steering angles for transmission or reception, superimpose them, broaden the band in the horizontal direction, and then perform super-resolution processing on them to enhance their effects. is. The superposition of plane waves etc. realizes the case where uniform focusing is performed regardless of the depth position with respect to the phase, and the apodization based on the weighting process of the above spectrum, for example, using a Gaussian function is performed. It is effective to compensate for the spectral intensity by targeting high-resolution imaging using aperture plane synthesis based on apodization using square waves or power functions. In other words, the superposition corresponds to calculating the inverse of the calculation of the Angular spectrum based on plane wave separation in Fourier imaging. Images that have undergone focus or aperture synthesis may be superimposed. It is important to adjust the steering angle to be superimposed, and if the angle difference is small, the spectral intensity is not broadened just by looking at the spectral intensity relatively in the lateral direction. Although there is an SN ratio in the same band (especially at the time of aperture plane synthesis processing), if a certain degree of angle difference is provided and overlapped, the band will be widened in a situation where there is spectral intensity. For a band with relatively low spectral intensity when the angle difference is small, for example, the spectral weighting process is effective. Since the signals are superimposed with different phases, the intensity of the signal is low, so it is necessary to perform double-precision calculation processing (beam forming, super-resolution processing, etc.) as necessary. On the other hand, when there is a certain degree of angular difference, a broadband signal can be easily obtained with a small number of waves or beams. In them, it is also effective to normalize and superimpose the wave energy. Similarly, the other two super-resolutions are also effective. Of course, superimposition processing in which waves with different focus positions and ultrasonic frequencies are superimposed is also effective as a super-resolution method. It is also possible to superimpose the results obtained by applying super-resolution to each wave, but superimposing and processing requires less processing and is more effective. Phase compensation (compensating for inhomogeneity of wave propagation velocity) in superposition is important.
The present inventors have so far used the cross-spectral phase gradient method, the multidimensional autocorrelation method, etc. as methods for observing the displacement vector (strain tensor) of the observation target based on the ultrasonic echo method (that is, the reflection method) and the transmission method. reported (and many others). It can also be used when waves other than ultrasonic waves are used. To improve the measurement accuracy when using these techniques, it is useful to extend the transverse modulation method to use more waves or to use spectral frequency division to achieve over-determined systems. , and regularization (a priori or a posteriori, cross-validation method) and weighted least squares (a priori or a posteriori) are effective. Maximum likelihood estimation (for example, Non-Patent Document 37, etc.) is also effective (with or without MAP). It is also effective to fuse, integrate, and mix them. When using variation, a locally stationary process can be assumed, Cramer-Rao Lower Bound (CRLB), Ziv-Zakai Lower Bound (ZZLB), etc. can be used. We have found that the variability estimated by Furthermore, by applying these displacement vector (strain tensor) observations, it is possible to improve the resolution of images (for example, ultrasonic echo images, etc.) obtained from wave signals obtained by the reflection method or the transmission method. For example, a plurality of temporally consecutive wave signal frames (e.g., ultrasound rf echo data frames) and image frames obtained therefrom are subjected to phase matching performed in the present invention, and motion based on a Markov model is performed. Using the prediction of the present invention, applying signal separation such as superimposition processing and ICA processing of the present invention, stochastic fusion (multiplexing and separation, etc.), various super-resolution (for example, non- Patent document 38 etc.) can be applied effectively. Fusion processing after super-resolution is also possible. Block matching is effective as the phase matching performed in these methods, but as described above, highly accurate phase matching based on phase rotation is particularly effective.
Also, so-called compressed sensing may be performed. Similarly, it may be done in DAS processing. Similar to the above three super-resolution, it may be applied to the superposition of multiple waves. Although the above three super-resolution methods are less computationally intensive, combinations of them are sometimes processed, including compressed sensing.

Next , DAS processing that may be implemented in the present invention will be described . This DAS processing also includes method D2, method DII, etc. of the present invention. In addition, although processing of real-time signals and analytic signals is described below, such processing may be applied to signals that have undergone various other processing described in this specification.
・DAS processing (Method D1) performed in a normal digital diagnostic device
In order to perform reception dynamic focusing, readout of reception signals stored in the memory of each channel after AD (Analog-to-Digital) conversion processing is determined by the distance between each receiving element position of each channel and each point of interest. The stored signal is read out from the memory whose address means the time of digital reception (that is, Delay). Then, those signals within the effective aperture width are added (summation). According to this method, an error determined by the sampling frequency of the received signal occurs in the delay. Therefore, although sampling is naturally performed based on the Nyquist theorem, sampling is performed at a frequency as high as possible. It is characterized by high speed. The signal to be processed may be an analytic signal in addition to the digital real signal after AD conversion processing. In addition, calculations may be performed using a general-purpose CPU as well as a dedicated circuit, and an example of a calculation protocol will be shown below. The sampled time and space coordinates (integer values) can be considered as the index of the array in which the digital signal is stored, or as the address of the memory in which the digital signal is stored (the same applies hereinafter). ).
For example, in DAS processing at the point of interest A or the sampling position i=I nearest to the point of interest A, represented by an integer value, the real or analytic signal
r(I) (DAS1)
If the signal from the point of interest A received by another receiving element is added to the received signal r' at a position longer by Δx in the propagation distance (in the case of the reflection method, aperture synthesis, etc., the forward path Or when performing beamforming for transmission, including the difference in propagation distance at that time), and when the sampling interval is expressed as sampx in distance, Δi = nint (Δx / sampx) (however, nint ( x) represents a rounded integer function), or Δi = inta (Δx/sampx) (where inta(x) represents a round-up integer function), or Δi = intd (Δx/sampx) (where intd(x) represents a rounding down integer function), a real signal or an analytic signal expressed using an integer value Δi calculated using a function or calculation that integerizes the argument x
r'(I+Δi) (DAS1')
is added. The integer value Δi is desirably such that Δi×sampx is closest to the analog value Δx (nint is preferable among the above), and the calculation method is not limited to these. The digital signals received by the receiving elements within the effective aperture width at each position are subjected to this processing and added. Under the assumption that the propagation speed is constant, a similar explanation can be given using time instead of distance.
Alternatively, there are various interpolation processing methods for digital signal values, and using them, the signal value at the position of Δx may be obtained by interpolation approximation (bilinear interpolation, high-order interpolation, Lagrangian method, spline interpolation, etc.). , and various interpolation approximations can be implemented).
・DAS processing (Method D2) with improved accuracy of Method D1
While performing DAS processing based on method D1, in order to improve the accuracy of the delay processing, the analytic signal a(t) of the received signal calculated based on the Hilbert transform is multiplied by a complex exponential function, and phase rotation is performed. By multiplying by Delay, the accuracy of t ₀ is obtained within a time shorter than the sampling time interval sapt, and addition processing is performed.
a(t+ _t0 )=a(t)exp[ _jω0 (t) _t0 ] (DAS2)
where j is an imaginary unit, the relationship t=i×sampt (i is an integer value from 0 to N-1) holds, and ω ₀ (t) is the nominal angular frequency at sampling time (position) t , centroid (center) angular frequency, and instantaneous angular frequency. Equation (DAS2) approximates the signal at the time (position) shifted by an analog amount _t0 in the positive direction of the sampling coordinate system with respect to the sampling time (position) t, but the point of interest A, or , the DAS processing at the sampling position i=I, which is expressed by an integer value and is closest to the point of interest A, similarly to method D1, the delay value Δt of the applied analog quantity (the reflection method In this case, when performing forward or transmission beamforming such as aperture synthesis, Δi (for example, Δi = nint (Δt / sampt)) calculated based on the delay at that time) The analytic signal a(t) (=a((I+Δi)×sampt)) of the represented sampling time (position) I+Δi is preferably used. That is, it is desirable to use the signal with the sampling time (position) t closest to the ideal analog time (position) T=I×sampt+Δt. Therefore, when T>t, the equation uses the positive value t ₀ =Tt=Δt−Δi×sampt, and when T<t, the equation uses the negative value t ₀ =Tt. The digital signals received by the receiving elements within the effective aperture width at each position are subjected to this processing and added.
Although the accuracy is higher than the method D1, it is strictly an approximation using the sampled time (position) frequency ω ₀ (t). Affected by frequency modulation such as attenuation and scattering. Similar to method D1, the higher the sampling frequency, the better. It has high speed.
By the way, when the sampling signal is clearly represented by discrete positions x (=i×sampx, i is an integer value from 0 to N-1) instead of time, the formula (DAS2) is expressed by wave number k ₀ (=ω ₀ /c = 2πf ₀ /c = 2π/λ) (where c is the propagation speed of the wave, f ₀ is the nominal frequency, the center of gravity (center) frequency, or the instantaneous frequency, λ is the wavelength) using
a( _x +x0)=a(x)exp[ _jk0 ( _x )x0] (DAS2')
is represented. Similarly, in the DAS processing at the point of interest A or at the sampling position i=I, which is represented by an integer value and is closest to the point of interest A, the analog distance difference Δx (in the case of the reflection method, When performing forward path or transmission beamforming such as aperture synthesis, use Δi (for example, Δi = nint (Δx / sampx)) calculated based on the difference in propagation distance at that time) Preferably, the analytic signal a(x) at the sampling position I+Δi (=a((I+Δi)×sampx)) is used. That is, it is desirable to use the signal at the sampling position x that is closest to the ideal analog position X=I*sampx+Δx. Therefore, when X>x, the equation uses a positive value x ₀ =X-x=Δx-Δi×sampx, and when X<x, the equation uses a negative value x ₀ =X-x.
Alternatively, in each of the equations (DAS2) and (DAS2'), the phase rotation is applied to the digital signal at the sampling time (position) t and position x closest to the ideal analog time (position) T and ideal analog position X. Instead of multiplying, each receiving element position within the effective aperture width when considering each sampling position i (i representing x=i×sampx) of each beam or wave to be generated as each point of interest A is Using the delay Δt and the distance difference Δx with respect to point A as discrete (digital) data Δt(i) and Δx(i) represented for each point of interest i, reception within the effective aperture width of each position Multiply the analytic signal a(i) of each digital received signal received by the element by a complex exponential function as follows,
a(i)exp[ _jω0 (i)Δt(i)] (DAS2'')
a(i)exp[ _jk0 (i)Δx(i)] (DAS2''')
to add.
・DAS processing (Method DII), which is an improvement of Method D2
Method D2 is an approximation process using the frequencies and wavenumbers of the sampling positions in the analytic signal, but the method DII of the present invention enables high-speed calculations without using the frequencies and wavenumbers of the sampling positions.
Centroid (central) frequency (analog value obtained at frequency coordinate i represented by an integer value in the discrete Fourier transform) that can also be obtained as the centroid of the amplitude spectrum S(i) (spectrum of non-positive frequencies is zero) in the frequency domain
M0 ₌ ΣiS(i)/ΣS(i) (DASII1)
However, the amplitude spectrum S(i) (where i=0 to N-1) is the product of the Fourier transform of the digital spatial signal r(x) (where the sampling position is x=i×sampx) and its conjugate. Calculated by square root. The addition range of Σ in the formula is i=0 to N/2.
Therefore, the wavenumber k ₀ is
k ₀ =(2πM ₀ )/(N×sampx) (DASII2)
is. where the analytic signal a(x) is
a(x)=A(x)exp{jk ₀ (x)x}
=A(x)exp{ _jk0 (x)×(i×sampx)} (DASII3)
I will try to express it.
Based on this, each receiving element position within the effective aperture width when each sampling position i (i representing x = i × sampx) of each beam or wave to be generated is considered as each interest point A is each interest point Using the distance difference Δx with respect to A as discrete (digital) data Δx(i) represented for each point of interest i, Multiply the analytic signal a(i) by a complex exponential function as follows,
a(i)exp{j(2πΔx(i)/sampx)/(N×sampx)×(i×sampx)}
=a(i)exp{j(2πΔx(i))/(N×sampx)×i} (DASII4)
and calculate. As the DAS-processed signal advances in the positive direction of x, an effect is obtained in which the direction orthogonal thereto is strongly broadened (higher resolution). Note that i (= 0 to N) in formula (DASII4) is instead
Ni (i=0 to N-1) (DASII5)
can be calculated using In that case, as the DAS-processed signal returns to the negative direction of x, the direction orthogonal thereto is strongly broadened (higher resolution). If Δx is constant independent of the interest point i, it means applying an invariant frequency modulation in the i direction. In other words, this processing applies frequency modulation to the received signal at each receiving element position within the effective aperture width for each point of interest i.
M0 + Δx(i)/ _sampx (DASII6)
It is a process of multiplying and adding. If the signal is represented as a digital time signal r(t) instead of a digital spatial signal, similarly for its analytic signal a(i) represented using the instantaneous frequency ω ₀ (i): is multiplied by a complex exponential function,
a(i)exp{j(2πΔt(i)/sampt)/(N×sampt)×(i×sampt)}
=a(i)exp{j(2πΔt(i))/(N×sampt)×i} (DASII4')
and calculate. The same is true for others.
It is also effective to apply a desired frequency modulation to a specific position or apply spatially uniform frequency modulation. If k ₀ ' is the modulation wavelength applied to the signal at position x and time t, and ω ₀ ' is the modulation frequency, then
a'(x)=a(x)exp[jk ₀ '(x)x] (DASII7)
a'(t)=a(t)exp[ _jω0 '(t)t] (DASII7')
Asked. It can be applied to discrete signals as well.
Superimposing unmodulated waves and modulated waves is also effective for broadening the bandwidth (increasing resolution) and increasing the accuracy of phase-based observations (displacement observations, etc.).
・Theoretically the most accurate DAS processing (Method D3)
In the method invented by the present inventor in the past (Patent Document 6, Non-Patent Document 15), the spectrum A(ω) of the local signal including the signal at the position of the interest point is multiplied by a complex exponential function in the frequency domain, and the local Delay by rotating the phase of the signal.
A'(ω)=A(ω)exp[jωt ₀ ]
This is an interpolation process that satisfies the sampling theorem and is theoretically the most accurate, but requires a long calculation time.
・Fourier beamforming (Method D4)
It is the beamforming that is the basis of the present invention. This method performs digital wavenumber mapping in the multidimensional frequency domain of the received signal, performs fast Fourier transform, and can perform calculations with an accuracy equivalent to that of method D3 at much higher speed. As a feature different from other reported ordinary Fourier imaging methods, the digital wavenumber mapping does not require interpolation approximation processing, but it is possible to perform interpolation approximation processing in order to speed up the mapping. However, in that case, it is required to increase the sampling frequency in order to reduce the occurrence of artifacts and the decrease in accuracy.
In these beamforming processes, the analytic signal is not necessarily processed, and the calculation time may be shortened by omitting the calculation for converting the received signal into the analytic signal. However, in the process of multiplying the complex exponential function in space and time, although it is theoretically incorrect and contains errors, imaging of the ultrasonic signal itself and various imaging (elasticity imaging, etc.) and many others) may be practical.
In addition, each delay processing in these DAS processing (methods D1 to D3) is not only in DAS processing, but also in other beamforming such as Fourier beamforming, adaptive beamforming, minimum variance, etc., and the signal described in this application It is useful for shifting signals for phase aberration correction, motion compensation, phase matching, alignment, position correction, etc. in terms of time coordinates, space coordinates, and spatio-temporal coordinates. Moreover, it is not limited to what is described in the present application. As with one-dimensional analytic signals, multi-dimensional analytic signals at position (x,y,z) or time (t1,t2,t3) can also be processed. For example, in method D2,
In the two-dimensional case,
a(t1+t1 ₀ , t2+t2 ₀ )=a(t1,t2)exp[j{ω1 ₀ (t1,t2)t1 ₀ +ω2 ₀ (t1,t2)t2 ₀ }]
(2 DAS2)
a( _x +x0, y+y0) ₌ a(x,y)exp[j{ _kx0 (x,y)x0 + _ky0 ( _x ,y)y0 _} ] (2DAS2')
In the case of three dimensions,
a(t1+t1 ₀ , t2+t2 ₀ , t3+t3 ₀ )=a(t1,t2,t3)
×exp[j{ω1 ₀ (t1,t2,t3)t1 ₀ +ω2 ₀ (t1,t2,t3)t2 ₀ +ω3 ₀ (t1,t2,t3)t3 ₀ }]
(3DAS2)
a(x+ _x0 , y+ _y0 , z+ _z0 )=a(x,y,z)
×exp[j{ _kx0 (x,y,z)x0+ _ky0 (x,y,z)y0 _{+ kz0} ( _x ,y,z) _z0 }] (3DAS2')
can be calculated as However, (ω1 ₀ , ω2 ₀ , ω3 ₀ ) are the nominal angular frequency, centroid (center) angular frequency, and instantaneous angular frequency in each time direction at each time (t1, t2, t3), and (kx ₀ , ky ₀ , kz ₀ ) are the wave numbers in each direction at each position (x, y, z). Shifting can be done in the same way for multidimensional signals of combinations of positions (x, y, z) or times (t1, t2, t3). In other words, in the digital signal, discrete coordinates (requiring high precision A high sampling frequency is desirable in some cases), analog shifting can be applied. As in the one-dimensional case, it can be more accurate than method D1 (which corresponds to block matching or fragment matching of the signal in the spatio-temporal domain), and the other method (phase matching based on phase rotation in the frequency domain of method D3 and Fourier beamforming of method D4), but the calculation speed is fast. Suitable for applications requiring high computational speed.
The modulation in method DII can also be applied to the multidimensional analytic signal at position (x,y,z) or time (t1,t2,t3) as for the one-dimensional analytic signal.
For example, in the two-dimensional case,
a'(x,y)=a(x,y)exp[j{kx ₀ '(x,y)x+ ky ₀ '(x,y)y}] (2DASII7)
a'(t1,t2)=a(t1,t2)exp[j{ω1 ₀ '(t1,t2)t1+ω2 ₀ '(t1,t2)t2}] (2DASII7')
In the case of three dimensions,
a'(x,y,z)=a(x,y,z)exp[j{kx ₀ '(x,y,z)x+ ky ₀ '(x,y,z)y+ kz ₀ '(x, y,z)z}]
(3DASII7)
a'(t1,t2,t3)=a(t1,t2,t3)exp[j{ω1 ₀ '(t1,t2,t3)t1+ω2 ₀ '(t1,t2,t3)t2+ω3 ₀ '( t1,t2,t3)t3}]
(3DASII7')
Asked. However, (kx ₀ ', ky ₀ ', kz ₀ ') are the modulation wavelengths in each direction applied to the multidimensional signal at each position (x, y, z), and (ω1 ₀ ', ω2 ₀ ', ω3 ₀ ') is the modulation frequency in each time direction applied to the multidimensional signal at each time (t1, t2, t3). Multidimensional signals with spatial and temporal coordinates can be similarly applied. Applications are not limited to DAS processing, including the one-dimensional case.
Other delay processing is also capable of multidimensional processing and can be applied to multidimensional signals.

In addition , in all beamforming such as Fourier beamforming and DAS processing described in (1) to (6) and (7) above, an array aperture element group (each element is independently driven and , having independent transmission or reception channels that can independently obtain received signals, the same transmission or reception to at least two or more elements in adjacent or separated positions (not necessarily equidistant) By applying a delay and apodization for transmission or reception and using it as one aperture, waves with greater strength than transmission or reception using one element may be used for beamforming for transmission or reception. For example, in a one-dimensional array transducer, if the element width and element spacing are shortened in order to improve the spatial resolution in the axial direction and lateral direction, the intensity of transmitted and received waves will be weakened. This is also the case when using a two-dimensional array or a higher-dimensional array (the element width and element spacing in the direction of the number of dimensions become shorter). Even when the thickness of the element is reduced to increase the frequency, or when using PVDF or the like, which has a lower transmission strength than PZT or the like in ultrasonic waves, the strength of the wave becomes weak. It is effective in such a case. That is, it is possible to increase the strength of the transmitted or received wave and improve the SN ratio of the received signal received by the receiving transducer. For example, it is effective when performing classical aperture plane synthesis (monostatic or multistatic). In this case, scanning can be performed by shifting the effective aperture width by a physical element interval (one element at a time), or by an arbitrary number of element intervals. Alternatively, the distance to be shifted between scans may be variable. Multi-dimensional array transducers may scan with different numbers of elements and different distance intervals in each dimension. The combination of adjacent or spaced-apart elements that are considered one aperture may be varied during scanning. Also, the effective aperture width may be changed during scanning, as is done in normal beamforming. It can be implemented in various types of array-type transducers, and the processing itself can be performed not only in digital devices, but also in analog devices and fusion devices with digital devices (e.g., transmission circuits including transmission delay, transmission apodization, etc. are analog circuits). It may also be implemented in In this process, the number of elements regarded as one element for transmission or reception, their positions (combinations), the effective aperture width for transmission or reception, and the distance to shift the effective aperture width for transmission or reception are added to the above beamforming parameters. , optionally also using the above wave parameters, and optionally using multiple effective apertures together to generate multiple waves or beams (simultaneous transmission or different targets in the same phase). time), and may similarly perform various applications described in the present invention. In transmission or reception in classical aperture plane synthesis or normal beamforming, multiple aperture elements regarded as one aperture may be used simultaneously at multiple locations within at least one effective aperture width. Physics When mechanically scanning (transmitting or receiving position changes digitally or analogously) an observation target with strong waves using a transducer with a wide aperture or a transducer with a single aperture, physical Beamforming may be applied to transmitted or received spatially dense or sparse relative to the width of the aperture.

In this simultaneous drive of multiple elements, even during reception by one element or multiple elements in single-element driving, the reception signals of those elements may be added to obtain the same effect. In addition, the same effect may be obtained by processing in the same manner at the time of reception by a single element or a plurality of elements in simultaneous drive of a plurality of elements. Also, when performing transmission beamforming by driving a group of elements within the effective aperture width, there are cases where multiple elements in various combinations within the effective aperture width transmit simultaneously. Transmission may be performed simultaneously from multiple effective aperture widths.

Simultaneous transmission of a plurality of elements in various manners as described above is equivalent to increasing the width of the transmitting elements and the pitch of the transmitting elements, resulting in grating lobes. As noted above, the width of the transmit elements may be large and the pitch of the transmit elements may be less than the element width. In some cases, the element width and pitch are physically large (in some cases, signals are collected more densely than the distance of the element width and pitch by mechanical scanning). These depend on the carrier frequency of the wave and the steering angle (including 0°) (geometric calculations can also be used to estimate the direction of the grating lobes). Moreover, side lobes are generated from the aperture element in the first place. Since these usually cause artifacts, beamforming is performed by optimizing the element width, pitch, element shape, wave carrier frequency, steering angle, apodization, etc., in order to reduce them. In practice, the sensitivities of transmission and reception are different, but they are often considered the same. Grating lobes and side lobes occur in reception alone as well, and can occur in both transmission and reception, but what is produced in transmission and reception can be different using different elements or combinations of elements. be.
As disclosed by the present inventor in Patent Document 7 and Non-Patent Document 19, it is possible to separate waves with different steering angles in the frequency domain (see also paragraph 0623). From the center of gravity of the spectrum and the instantaneous frequency of the analytic signal corresponding to the spectrum, the wave propagation direction and the actually generated steering angle can be estimated. This method can also be used to separate or remove these grating lobes and side lobes from the main lobe. Usually, the spectra of grating lobes and side lobes are separated in frequency space, and it is easy to identify and separate them (by extracting only the spectrum of interest or by zeroing the spectra of other bands). better).

In fact, these grating lobes and side lobes can be actively used as they are or modified to realize transverse modulation, and can be applied to various applications such as wave imaging and displacement (vector) observation. can. Transverse modulation usually refers to the crossed wave situation, but may also be applied separately to each steered wave (sometimes weighted at the same time). In these, for example, separated waves can be detected and converted into incoherent signals for imaging, or these incoherent signals can be superimposed (compounded) for imaging to reduce speckle. Alternatively, the spectrum of the coherent signal may be divided (or weighted) in certain bands to synthesize a new coherent signal. It is also possible to generate one or more waves with different parameters such as steering angles and detect and image each of them, or superimpose them for imaging. Also, each of the coherent signals generated by these processes can be used to observe displacement, or used simultaneously to solve simultaneous equations or over-determined systems for the same displacement or displacement vector components. Sometimes. Some of the waves corresponding to grating lobes and side lobes have steering angles larger than the steering angle set by beamforming. The measurement accuracy of lateral displacement (Non-Patent Document 19) and displacement vector is improved. There is a steering angle that is larger than or substantially the same as the steering angle set by beamforming, and it is sometimes used in combination with these. A plurality of grating lobes and side lobes are generated, but the higher the frequency in the horizontal direction, the weaker the signal strength. It is used as a component of the wave motion along with the low frequency one in the horizontal direction. As with normal steering beamforming, the higher the lateral frequency, the lower the frontal frequency. However, when observing the displacement vector, the effect of improving the measurement accuracy of lateral displacement outweighs the effect of lowering the frequency. , the measurement accuracy of the displacement in the front direction may also be improved.

Using a plurality of waves generated in these processes, a displacement vector that moves in an arbitrary direction (multidimensional autocorrelation method, multidimensional self-Doppler method, etc. (Non-Patent Document 13) is used in the digital signal unit. In the method of solving simultaneous equations related to the components, the present inventor's past invention) and ordinary displacement in one direction can be measured with high accuracy (deriving more equations than the number of displacement components to be obtained, In a (over-determined) system, the method of least squares, the average value of the calculation results, and the overlap of waves to obtain highly accurate results in a high frequency and wide band state, etc. Patent Document 5). One equation is derived from each generated wave. A single wave may be subjected to normal Doppler techniques. Each wave may be a superposition of waves, or the spectrum may be frequency-divided or spectrum-processed. Each wave preferably has a high frequency, may be used with the low frequency spectrum removed, and is preferably broadband if high spatial resolution is also required (non-patented Reference 14). The division and processing may use windows that allow the spectrum to be weighted. Strain (tensor), strain rate (tensor), velocity (vector), and acceleration (vector) are obtained from the displacement (vector) by partial differential processing using a spatial or temporal differential filter. These can be used to determine (visco)shear modulus, viscosity, mean normal stress, density, and the like. In addition, as a method of measuring displacement (vector), a multidimensional cross-spectrum phase gradient method (one of the block matching methods, Patent Document 6, Non-Patent Document 15, etc.), which is a past invention of the inventor of the present application, is used. See) and the digital demodulation method (Patent Document 7), and it is also possible to measure distortion and the like. They can also be used to measure the propagation of shear waves and low-frequency vibration waves. (Visco)shear modulus, shear wave propagation velocity, propagation direction, shear wave displacement, frequency, phase, vibration amplitude, vibration velocity, vibration acceleration, etc. can be measured. These can also be determined as distributions.

In order to improve the accuracy of this displacement measurement, the inventor of the present application invented regularization in the past. In order to determine the regularization parameter of the penalty term, for example, a posteriori, the variation of the measured displacement (vector) is estimated and used under a (local) stationary process (Patent Document 6), or a priori, Using the characteristics of waves or beams, etc., we have invented the use of Ziv-Zakai Lower Bound (ZZLB, for example, the variation expressed in Non-Patent Document 16) (for example, Non-Patent Documents 17 and 18). It is also possible to apply the variation derived by the one-dimensional autocorrelation method (Non-Patent Document 20), which is a velocity measurement method in one direction (used in power Doppler).

In the present invention, these variations and ZZLB can also be used for weighting in order to adjust the reliability of each equation when combining the Doppler equations derived as described above (high reliability is heavy , set low for those with low confidence). That is, at each location in the region of interest, obtain their values from each of the above waves or beams, and weight each of the Doppler equations derived from the corresponding waves or beams at each location to form a system of equations solve. Using the method of least squares, the Weighted Least Squares Solution (WLSQS) can be determined a posteriori or a priori.

但し、uは関心点又は関心点を含む局所領域の未知変位ベクトル又はその分布、bはフレーム間に生じた関心点又は関心点に関する局所領域の位相変化又はその分布を表すベクトル、Aは対応して並べられた各関心点又は各関心点に関する局所領域の周波数成分又はその分布からなるマトリクスであり、Aとbの成分は、時間方向や空間方向に移動平均処理されていることがある。また、デモジュレーションされている場合には、搬送周波数を持つ２方向又は１方向の変位成分のみを未知とする状態のドプラ方程式が連立された状態にある。
と表されるとき、ばらつきやＺＺＬＢの逆数を用いて、そのもの、又は、そのべき乗、又は、その分布を表すマトリクスWを用いて重み付けして解く。

The system of Doppler equations derived above is

However, u is the unknown displacement vector of the point of interest or the local area containing the point of interest or its distribution, b is the point of interest occurring between frames or the phase change of the local area about the point of interest or its distribution, and A is the corresponding It is a matrix consisting of frequency components or their distributions in local regions related to each interest point or each interest point arranged in a matrix, and the components A and b may be subjected to moving average processing in the temporal direction and the spatial direction. In the case of demodulation, Doppler equations in which only displacement components in two directions or one direction with a carrier frequency are unknown are in a simultaneous state.
is weighted and solved using the variation or the reciprocal of ZZLB, or its exponentiation, or the matrix W representing its distribution.

但し、Ａxp、Ａyp、Ａzp（p=１～Ｎ）は、ｘ、ｙ、ｚ方向の周波数成分であり、式（Ａ１）又は式（Ａ２）のマトリクスＡの成分であり、また、bp(p=１～Ｎ)は、フレーム間の位相変化であり、同式のベクトルbの成分であり、Ｗｐは式（Ａ２）のマトリクスＷの対角成分である。クロススペクトル位相勾配法（ブロックマッチング法の１つ）を用いる場合と、多次元自己相関法や多次元ドプラ法に基づいてブロックマッチングを行った場合とにおいては、連立された方程式の全てにＷｐが掛かる（即ち、１つのｐにおいて、複数の方程式が連立され、それらの全てにＷｐが掛かる）。 Specifically, focusing on one point of interest or local region, one (or more) Doppler equations derived from one of the waves or beams p (= 1 to N), i.e., the cross-spectral phase gradient method is used, the number of simultaneous equations that hold for the phase spectrum in the signal band for the cross spectrum obtained in the local area, and the block based on the multidimensional autocorrelation method and the multidimensional Doppler method In the case of matching, the number of simultaneous equations that hold at the position in the local region), the inverse value Wp of ZZLB calculated at the point of interest or the local region is the Since it is the reciprocal of ZZLB of the displacement in the beam direction, for example, if the unknown displacement of the point of interest or local region is the 3-dimensional vector u = (Ux,Uy,Uz) ^T ,

where Axp, Ayp, and Azp (p=1 to N) are frequency components in the x, y, and z directions, the components of matrix A in formula (A1) or formula (A2), and bp(p = 1 to N) are the phase changes between frames and are the components of the vector b in the equation, and Wp is the diagonal component of the matrix W in equation (A2). In the case of using the cross-spectral phase gradient method (one of the block matching methods) and in the case of performing block matching based on the multidimensional autocorrelation method or the multidimensional Doppler method, all of the simultaneous equations have Wp Multiply (ie, at one p, multiple equations are simultaneously multiplied, all of which are multiplied by Wp).

但し、Ｔは多次元自己相関法や多次元ドプラ法のときは周波数や位相変化を求める際の移動平均幅であり、多次元クロススペクトラム位相勾配法や多次元自己相関法、多次元ドプラ法にてブロックマッチングを行うときは変位計測の局所領域のビーム方向の長さであり（連立するビーム又は式においてＴが同一の場合には、変数Ｔを使用する必要はなく、１等の任意定数で良い）、また、f_0bはビーム方向の超音波周波数、Ｂ_ｂはビーム方向のＲｅｃｔａｎｇｕｌａｒ帯域幅であり、ＳＮＲcはエコーのＳＮ比であるＳＮＲeと相関ＳＮ比であるＳＮＲρ（変位や対象の変形に伴って波形が歪むことにより相関性が低下して生じるノイズ成分に対する信号比）

但し、ρはフレーム間で局所クロススペクトルを求めた際の相関値、又は、移動平均幅の長さにて局所に求められた相関値である。
との結合(Ｃｏｍｂｉｎｅｄ)ＳＮ比

である。 For example, according to ZZLB described in Non-Patent Document 16, when Cramer-Rao Lower Bound (CRLB) holds, its square variance is

However, T is the moving average width when obtaining frequency and phase changes in the case of the multidimensional autocorrelation method and the multidimensional Doppler method. When block matching is performed by using good), f _0b is the ultrasonic frequency in the beam direction, Bb is the rectangular bandwidth in the beam direction, and _SNRc is the echo SN ratio SNRe and the correlation SN ratio SNRρ (depending on displacement and deformation of the object). (Signal ratio to the noise component that is generated due to the deterioration of the correlation due to waveform distortion)

Here, ρ is a correlation value obtained when the local cross spectrum is obtained between frames, or a correlation value obtained locally based on the length of the moving average width.
Combined SN ratio with

is.

式（Ｓ１）及び式（Ｓ２）中のS(fb)は、式（Ｓ１）において生のスペクトルとfb = １を用いて計算される全エネルギーを用いて生のスペクトルを除し、全エネルギーが１となる様に正規化されたものである。若しくは、式（Ｓ１）及び式（Ｓ２）において、S(fb)として生のスペクトルを用い、各々の計算結果を上記の全エネルギーで除してもよく、後者の方が計算量が少ない。式（Ｓ２）で求まるＢ_bは、実際の送信又は受信したパルス形状又はスペクトル形状を基にＲｅｃｔａｎｇｕｌａｒ帯域幅を表す様に換算して使用されることもある（以下の第２次の中心モーメントを用いる場合も同様である）。多次元受信信号の場合において、ビーム方向の周波数軸と直交する２軸（即ち、３次元）又は１軸方向（２次元）も含めて計算されることもあり、例えば、受信信号が３次元の場合には、次式が用いられる。

Therefore, the variation can be estimated by measuring and using T, f _0b , B _b , SNRc, SNRe, SNRρ, ρ as described in Patent Document 17, for example. Arbitrary constants and typical values may be used without measurements. As described in Non-Patent Document 19, f _0b can be obtained by obtaining the instantaneous frequency or the first moment (the center of gravity, that is, the weighted average value), and B _b is the second central moment It can be estimated by taking the square root.

S(fb) in equations (S1) and (S2) divides the raw spectrum using the total energy calculated using the raw spectrum and fb = 1 in equation (S1), and the total energy is It is normalized to be 1. Alternatively, in equations (S1) and (S2), the raw spectrum may be used as S(fb) and each calculation result may be divided by the above total energy, the latter being less computational. B _b determined by the formula (S2) may be converted to represent a rectangular bandwidth based on the actual transmitted or received pulse shape or spectrum shape (the following second-order central moment is The same is true when using In the case of a multidimensional received signal, the calculation may include two axes (that is, three dimensions) or one axis (two dimensions) orthogonal to the frequency axis of the beam direction. , the following formula is used:

Similarly, the spectra in formulas (S1′) and (S2′) are normalized to have unity energy. Alternatively, similarly, the raw spectra may be used in equations (S1′) and (S2′) and each calculated result divided by the total energy, the latter being less computationally demanding. B _b obtained by the formula (S2′) may be converted to represent the rectangular bandwidth based on the actual transmitted or received pulse shape or spectrum shape (the following second central moment is used).
On the other hand, the echo signal-to-noise ratio, SNRe, can be statistically estimated by repeatedly collecting echoes for an object or calibration phantom. In some cases, a typical value is used a priori and determined based on the object, its state, experience of measurement, and the like. On the other hand, the correlation signal-to-noise ratio SNR ρ can be estimated using the correlation value ρ that is evaluated locally at each point of interest. These methods of obtaining are not limited to these. Also, if any value is missing and the variation cannot be estimated absolutely, a typical value can be used, but when setting the regularization parameter, it is expressed within the possible range. It is possible to obtain the best results from the results obtained while changing this constant of proportionality by multiplying the variability obtained by an undetermined constant of proportionality (Regarding regularization, for example, Patent Document 6, Non-Patent Document 17, 18).

と表されるとき、ドプラ角周波数ωの平均と分散は、各々、

と表され、パルス繰り返し周期Ｉ内でビーム方向の速度が一定であるとすると、近似的に

パルス繰り返し周期Ｉが短いとすると、近似的に

と求められる。ちなみに、Ｒ(０)はWiener-Khinchine theoremに基づいて、信号のエネルギー又はパワースペクトルの積分から求めても良い。
尚、上記の自己相関関数は空間座標を用いて表さなかったが、関心点におけるビーム方向の１次元局所領域や関心点を含む２次元局所領域又は３次元局所領域の信号を用いて処理しても良い（１次元や２次元又は３次元の移動平均処理を含めても良く、例えば、非特許文献１３や１９に詳しい）。その２次元又は３次元の局所領域において、ビーム方向がそれらの局所領域の直交座標軸（デカルト座標系だけで無く、曲線座標系を含む）である必要は無く、求まる平均とばらつきは、その局所領域のビーム方向の速度の平均とばらつきである。
従って、ビーム方向の変位の平均と分散の各々は、式（ＡＵＴＯ２）又は（ＡＵＴＯ２'）と（ＡＵＴＯ３）又は（ＡＵＴＯ３'）のそれぞれに、ＩとＩの二乗を掛けて求められる。ここでは、変位のばらつきを用いた変位の重み付き計測を示したが、速度や加速度のドプラ方程式を導出して各々のばらつきを重みにしてそれらを計測しても良い。 Also, when applying the variation (used in power Doppler) derived by the one-dimensional autocorrelation method (Non-Patent Document 20), which is a unidirectional velocity measurement method, the autocorrelation function is slow At -time-axis τ,

, the mean and variance of the Doppler angular frequency ω are, respectively,

Assuming that the velocity in the beam direction is constant within the pulse repetition period I, approximately

Assuming that the pulse repetition period I is short, approximately

is asked. Incidentally, R(0) may be obtained from integration of signal energy or power spectrum based on the Wiener-Khinchine theorem.
Although the above autocorrelation function was not expressed using spatial coordinates, it was processed using signals of a one-dimensional local area in the beam direction at the point of interest, a two-dimensional local area including the point of interest, or a three-dimensional local area. (One-dimensional, two-dimensional, or three-dimensional moving average processing may be included, for example, see Non-Patent Documents 13 and 19 for details). In the two-dimensional or three-dimensional local region, the beam directions need not be the Cartesian coordinate axes of those local regions (including not only Cartesian coordinate systems but also curvilinear coordinate systems), and the averages and variances found are the local region are the mean and variation of the velocity in the beam direction of .
Therefore, each of the mean and variance of the displacement in the beam direction is obtained by multiplying the equations (AUTO2) or (AUTO2') and (AUTO3) or (AUTO3') by I and I squared, respectively. Although weighted measurement of displacement using displacement variation is shown here, Doppler equations of velocity and acceleration may be derived and each variation may be weighted to measure them.

や、ＺＺＬＢとは別の方法が用いられる場合において、ビーム方向変位のばらつきが直接的に推定されずに変位ベクトルの各成分のばらつきが推定される場合には、以下の如く推定できる。つまり、それらの変位成分の計測誤差の確率過程が独立である仮定の下で、ビーム方向変位への誤差の伝搬を考えれば良い。例えば、３次元変位ベクトルの変位の平均値とばらつきの各々が、（ｍｘ，σｘ）、（ｍｙ，σｙ）、（ｍｚ，σｚ）と推定されたとき、ビーム方向の変位の平均値ｍ_beamとばらつきσ_beamの各々は、以下の如く推定できる。

Here, when the first-order moment and the second-order central moment in the beam direction are not directly estimated but are estimated in each direction (for example, when the signal is three-dimensional), the x-axis direction The first-order moment f _0x and the second-order central moment Bx are

Alternatively, when a method other than ZZLB is used and the dispersion of each component of the displacement vector is estimated instead of directly estimating the dispersion of the beam direction displacement, the estimation can be made as follows. In other words, under the assumption that the stochastic processes of the measurement errors of these displacement components are independent, the error propagation to the beam direction displacement should be considered. For example, when each of the average value and variation of the displacement of the three-dimensional displacement vector is estimated as (mx, σx), (my, σy), and (mz, σz), the average value of the displacement in the beam direction m _beam and Each of the variations σ _beam can be estimated as follows.

と推定できる。 Further, the parameters (T, f _0b , B _b , SNRc, SNRe, SNRρ) described in formulas (A4) to (A6) are given for each of a plurality of directions, and the average displacements f _0x and f _0y in each direction are given. , f _0z and the variations σ _CRLBx , σ _CRLBy , and σ _CRLBz are respectively estimated, the variation σ _CRLB of the displacement in the beam direction is obtained according to equation (A8) as follows:

can be estimated.

It is similarly derived when the unknown displacement of the position of interest is a two-dimensional vector u = (Ux, Uy) (when there is one unknown displacement and it is a displacement U in a certain direction or a displacement U in the beam direction uses the exact estimate obtained).

When obtaining the displacement at the point of interest or at each of the local regions related to the point of interest, the weighted Doppler equations (A3) (p=1 to N) at that position are combined to solve the equation (A2). The number of waves or beams (ie the number of equations) N must be greater than or equal to the number of unknown displacement components. However, when the above block matching is performed, a plurality of equations (A3) are established from one wave or beam p as described above. Therefore, measurements may be made with fewer waves or beams than when using other displacement measurement methods.

At the same time, when regularization is also applied, all the equations (A3) that are established at multiple points of interest in the region of interest or local regions related to the points of interest are simultaneous, and when the unknown vector u is a displacement component distribution (A2), and a Regularized Weighted Least Squares Solution (RWLSQS) may be obtained. Variation and ZZLB may be used as a regularization parameter at that time (a value proportional to variation, a value proportional to a power, etc.). Regularization is detailed in Patent Document 6, for example. The displacement variability in the direction of propagation of each individual wave or beam may be used as a regularization parameter for the displacement in all directions, and the variability in the displacement in each direction estimated for each wave or beam may be may also be used as a regularization parameter for displacements in the direction of . For example, as a distribution of an unknown three-dimensional displacement vector (Ux, Uy, Uz) ^T , an unknown vector whose partial vectors are Ux, Uy, and Uz, which are distributions of displacement components Ux, Uy, and Uz in the x, y, and z directions u=(Ux, Uy, Uz) When obtaining ^T , a matrix W having a diagonal component of the displacement variation Wp (p=1 to N) in the beam direction, or a displacement variation Wpx, Wpy in each direction, Using matrices Wx, Wy, and Wz having Wpz (p=1 to N) as diagonal components, the least-squared error energy E(u) and its solution u are expressed as follows.

Variation and ZZLB may be subjected to averaging of measurement results of displacement components obtained by simultaneous selected Doppler equations, and may be used for weighting in that case. Using the reciprocal of the displacement variation in the beam direction Wp (p=1 to N) or the reciprocal of the displacement variation in each direction (Wpx, Wpy, Wpz) (p=1 to N), the weighted average is may be requested.

In addition to the steady process and ZZLB, the variation may be obtained based on the ensemble average under a non-stationary process, may be obtained using a calibration phantom, or may be obtained from the measurement object itself. Regularization parameters and weighted matrices are determined in the manner described above, but there are also cases where typical values are a priori determined based on the subject, its state, measurement experience, and the like. Moreover, it is not limited to this.

３次元の場合には、例えば、ｘ＝ｒsinθcosφ、ｙ＝ｒcosθ及びｚ＝ｒsinθsinφのとき、ヤコビ行列は（２７）式で表されるフーリエ変換の場合の逆になるから（逆関数の定理）、

を得て（極座標系の原点は特異点）、関心点又は関心点を含む局所領域の信号とフレームNe内の同一位置の信号にて同様に直接に計算される極座標系のフーリエ変換（スペクトル）から求まる局所クロススペクトルの位相の勾配（クロススペクトル位相勾配法、非特許文献１５）や多次元自己相関法等の変位ベクトル計測法から推定される動径方向と極角、仰角、方位角の変位ベクトル成分を用いて極座標系のフーリエ空間で其のフレームNe内の探索領域又は関心領域の信号の極座標系のフーリエ変換（スペクトル）に複素指数関数を乗ずる位相回転を施すことにより、動き補償を施して位相マッチングすれば良い。例えば、２次元の場合に局所信号の変位ベクトルが(Δ_r,Δ_θ)と推定された場合には、探索領域の波動信号の２次元スペクトルF(k_r,k_θ)に複素指数関数exp{i(k_rΔ_r+ k_θΔ_θ)}を乗じて探索領域を変位ベクトルと逆の方向に(－Δ_r,－Δ_θ)だけ移動させ、３次元の場合に局所信号の変位ベクトルが(Δ_r,Δ_θ,Δ_φ)と推定された場合には、探索領域の波動信号の３次元のスペクトルF(k_r,k_θ,k_φ)に複素指数関数exp{i(k_rΔ_r+ k_θΔ_θ+k_φΔ_φ)}を乗じて探索領域を変位ベクトルと逆の方向に(－Δ_r,－Δ_θ,－Δ_φ)だけ移動させる。つまり、これを逆フーリエ変換したものから同一位置にて位相マッチングされたデカルト座標系又は極座標系で表された局所信号が得られる。即ち、２次元の場合には、k_ｘ＝k_ｒsinθ'及びk_ｙ＝k_ｒcosθ'であるから、各々、

であり、３次元の場合には、k_ｘ＝k_ｒsinθ'cosφ'、k_ｙ＝k_ｒcosθ'及びk_ｚ＝k_ｒsinθ'sinφ'であるから、各々、

である。
局所領域や探索領域は、関心点を中央にて設定することが多いが、後者は、組織変位の方向が先験的に与えられる場合に効率良くその方向に探索領域を設けることができる。位相収差補正や動き補償を目的とする場合も同様である。クロススペクトル位相勾配法以外の、例えば、多次元自己相関法等を用いる場合には、各フレームの（解析）信号を極座標系で求めれば良く、同様にヤコビ演算を用いたフーリエ変換を実施すれば良い。この様にして回転の位相マッチングも並進の位相マッチングと同様に高精度に実施できる。上記のヤコビ演算を用いたフーリエ変換により、セクタスキャンやコンベックス型探触子を用いた場合等、任意の直交座標系において表されるデジタル信号や離散フーリエ変換（離散スペクトル）を近似処理せずに直接に別の任意の直交座標系（直交デカルト座標系や様々な曲線直交座標系等の異なる直交座標系であったり、原点が異なったり回転していたり、同一の直交座標系であっても原点位置が異なったり、回転していたりする場合を含む）にて表し直すことが可能である（段落００２６や段落０１３２、０２１４等に記載の位相回転を用いた厳密な補間近似も精度が高いが、計算に時間を要する）。図２０は、ヤコビ演算を用いたフーリエ変換による信号処理の一例を示すフローチャートであるが、上記の信号処理はこの限りでは無い。
位相マッチングは、各直交座標系で表される信号に対し、まずは並進の位相マッチングを行った上で、回転の位相マッチングを行うことが多いが、その限りでは無い。交互に実施することもあるし、逆の順に実施することもある。また、非特許文献１３や１５にある通り、各々の位相マッチングを繰り返し実施することがある（修正変位量が予め設定した閾値よりも小さくなったら繰り返しを終了する）。伸縮の処理を交えることもある。
尚、局所領域や探索領域は必ずしも矩形では無く、円形等の他の形をしていることがある（データが矩形状の配列、例えば、正方形や長方形、立方体や直方体の配列に格納されている場合には、実寸の円形や球等の領域外に該当する位置の配列には零詰めされることがある）。周波数領域における複素指数関数を乗ずることによる位相回転によって探索領域内にて巡回した信号が関心領域内に現れない様に、探索領域は適切に局所領域よりも大きい必要がある（闇雲に大きくすると計算量が増えるだけであり、先験的に観測対象の変位ベクトルの大きさから適切に決める）。ちなみに、動径方向の変位が無く、観測対象が回転方向の変位のみの場合には、探索領域の大きさは局所領域と同一で良い。また、探索領域が前のフレーム内に設けられることもある。また、処理時間を短縮化するべく、複素指数関数を乗ずる位相マッチングでは無く、デジタル信号の離散的なシフティングによって位相マッチングを行うこともある。この場合には、必ずしも探索領域に動き補償を施す必要は無く、局所領域を別のフレームの探索領域内にて直接に探索（ブロックマッチング）して位相マッチングを行っても良い。
尚、位相マッチングは、極座標系の周波数領域における上記の位相回転処理の他に、段落０３８４に記載のＤＡＳ処理（方法Ｄ１の補間処理又は方法Ｄ２の時空間領域において解析信号に位相回転を施す処理を通じたシフティング）のＤｅｌａｙ処理を基礎として多次元信号をシフティングする場合と同様に、極座標系で表される多次元解析信号に対して処理して位相マッチング（動径方向のシフティングや回転処理）できる。上記の極座標系の周波数領域における位相回転処理は、方法Ｄ３のＤｅｌａｙ処理に基づく。これらの処理は、極座標系におけるＤＡＳ処理にも使用でき、極座標系におけるフーリエビームフォーミングやアダプティブビームフォーミングやミニマムバリアンス等の他のビームフォーミング時にも実施でき、また、本願に記載の信号の位相収差補正や動き補償、位相マッチング（新しい処理方法を含む）、位置合わせ、位置補正等の信号を、時間座標や空間座標や時空間座標において動径方向のシフティングや回転する場合に有用である。
例えば、２次元の場合には、
f(r+r₀, θ+θ₀)=f(r,θ)exp[j{k_r(r,θ)r₀+k_θ(r,θ)θ₀}] （２ＤＡＳｒ）
３次元の場合には、
f(r+r₀, θ+θ₀, φ+φ₀)=ｆ(r,θ,φ)
×exp[j{k_r(r,θ,φ)r₀+k_θ(r,θ,φ)θ₀+k_φ(r,θ,φ)φ₀}] （３ＤＡＳｒ）
と計算できる。但し、(k_r,k_θ,k_φ)は各位置(r,θ,φ)における各方向の波数である。つまり、デジタル信号において、各位置における各方向の波数を用いて、離散極座標系（高い精度を必要とする場合は高サンプリング周波数であることが望ましい）においてアナログ的なシフティングを施すことができる。１次元の場合と同様に、方法Ｄ１（時空間領域における信号のブロックマッチング又は断片のマッチングに該当）に比べて精度を高くでき、その他の方法（方法Ｄ３や方法Ｄ４）に比べると精度が低いが計算速度は速い。高速な計算速度を必要とする場合に適している。
尚、これらの処理は、観測対象そのものと共に、点拡がり関数も同時に動径方向や回転方向にシフティングさせることになることに注意が必要である（位相マッチングの誤差源となり得る）。また、これらの処理は、キャリア周波数を持たない信号において実施されることもある（例えば、画像処理等）。 In this way, weight values and regularization parameters can be set with high accuracy while maintaining spatial resolution. For example, the entire region of interest or, for example, a partial region within the region of interest, such as a propagation direction distance or a partial region provided for each depth in the observation target), estimates the variation, and for each wave or beam It may also be set and processed globally. In order to make the measurement itself possible and to improve the measurement accuracy, the phase matching method invented in the past by the inventor of the present application (Patent Document 6, Non-Patent Document 15) is required. Other reported stretching methods are useful for the conversion.
FIG. 19 shows motion compensation (phase matching) performed by moving a search area provided in the next frame to be processed in the two-dimensional case using the point of interest or an estimated displacement vector of a local area containing the point of interest. ), where FIG. 19(a) shows translational motion compensation and FIG. 19(b) shows motion compensation including rotation. The three-dimensional case is similarly processed using three-dimensional or two-dimensional local regions and search regions in a three-dimensional space.
When performing translation processing in phase matching (Non-Patent Documents 13 and 15), for example, in an arbitrary orthogonal coordinate system such as a Cartesian coordinate system using a linear array probe, each interest point or each interest point For each local region containing . The spectrum is multiplied by the complex exponential function in the same manner as the phase rotation of the one-dimensional signal in method D3 of paragraph 0384, for example, the displacement vector of the local signal is (Δ ₁ , Δ ₂ ) (1 and 2 are two-dimensional represents the axes of a Cartesian coordinate system), then the two-dimensional spectrum A(k ₁ ,k ₂ ) of the wave signal in the search area has the complex exponential function exp{i(k ₁ Δ ₁ +k ₂ Δ ₂ )} to shift the search area by (−Δ ₁ ,−Δ ₂ ) in the direction opposite to the displacement vector, so that the displacement vector of the local signal is (Δ ₁ ,Δ ₂ ,Δ ₃ ) in the case of three dimensions. ( ₁ , ₂ and 3 represent the axes of a _three -dimensional Cartesian coordinate system), then the complex Multiply the exponential function exp{i(k ₁ Δ ₁ +k ₂ Δ ₂ +k ₃ Δ ₃ )} to shift the search area by (-Δ ₁ ,-Δ ₂ ,-Δ ₃ ) in the direction opposite to the displacement vector (Patent Document 6, Non-Patent Document 15, etc.). That is, when the inverse Fourier transform is applied to the multiplied spectrum, a local (spatio-) spatial signal expressed in a Cartesian coordinate system phase-matched (shifted) at the same position is obtained. Incidentally, when shifting the local region in the direction of the displacement vector, exp{-i(k ₁ Δ ₁ +k ₂ Δ ₂ )} or exp{-i (k ₁ Δ ₁ +k ₂ Δ ₂ +k ₃ Δ ₃ )} can be multiplied, but in digital signal processing, as will be described later, the cyclic nature of the digital signal is a problem, so the wave signal in the search area should be shifted.
When performing rotation processing in phase matching, similar to translational phase matching (see Non-Patent Documents 13 and 15, FIG. 19A), for example, a Cartesian coordinate system using a linear array probe For each point of interest or each local region containing each point of interest in any Cartesian coordinate system such as The Fourier transform (i.e., the spectrum) of the polar coordinate system (the center of the polar coordinate system is each point of interest or location within or outside the region of interest) of the signal of the region or region of interest (see FIG. 19(b)) is described herein. It is directly obtained without approximation through the Jacobian operation implemented in the book.
That is, in the two-dimensional case, for example, when x = rsin θ and y = rcos θ, the Jacobian matrix is the inverse of the Fourier transform represented by equation (22) (inverse function theorem).

In the case of three dimensions, for example, when x = rsin θcos φ, y = rcos θ, and z = rsin θsin φ, the Jacobian matrix is the inverse of the Fourier transform represented by equation (27) (inverse function theorem).

(the origin of the polar coordinate system is the singular point), and the Fourier transform (spectrum) Radial direction, polar angle, elevation angle, and azimuth angle estimated from displacement vector measurement methods such as local cross spectrum phase gradient method (cross spectrum phase gradient method, Non-Patent Document 15) and multidimensional autocorrelation method Motion compensation is performed by applying a phase rotation that multiplies a complex exponential function to the polar coordinate system Fourier transform (spectrum) of the signal in the search area or the region of interest in the frame Ne in the polar coordinate system Fourier space using the vector component. phase matching. For example, when the displacement vector of the local signal is estimated to be (Δ _r ,Δ _θ ) in the two-dimensional _{case ,} the complex exponential function exp By multiplying {i(k _r Δ _r + k _θ Δ _θ )}, the search area is moved by (−Δ _r ,−Δ _θ ) in the direction opposite to the displacement vector, and the displacement vector of the local signal is is estimated to be (Δ _r ,Δ _θ ,Δ _φ ), the complex exponential _{function exp} { _i (k _r Δ _r + k _θ Δ _θ +k _φ Δ _φ )} to move the search area by (−Δ _r ,−Δ _θ ,−Δ _φ ) in the direction opposite to the displacement vector. In other words, the inverse Fourier transform of this provides a local signal expressed in a Cartesian coordinate system or a polar coordinate system phase-matched at the same position. That is, in the two-dimensional case, k _x =k _r sin θ' and k _y =k _r cos θ', so that

and in the three-dimensional case, k _x =k _r sin θ'cos φ', k _y =k _r cos θ' and k _z =k _r sin θ'sin φ', so that

is.
Local regions and search regions are often set with the point of interest at the center, but the latter can efficiently set the search region in the direction of tissue displacement when the direction is given a priori. The same is true when aiming at phase aberration correction or motion compensation. In the case of using, for example, a multidimensional autocorrelation method other than the cross-spectrum phase gradient method, the (analytical) signal of each frame may be obtained in a polar coordinate system, and a Fourier transform using the Jacobi operation may be performed. good. In this manner, rotational phase matching can be performed with high accuracy as well as translational phase matching. Without approximating digital signals and discrete Fourier transforms (discrete spectra) expressed in any orthogonal coordinate system, such as when sector scans and convex probes are used, by Fourier transform using the above Jacobian operation Any other Cartesian coordinate system directly (different Cartesian coordinate systems, such as Cartesian Cartesian or various curvilinear Cartesian coordinate systems, different or rotated origins, or even the same Cartesian coordinate system) (including cases where the position is different or rotated) can be re-expressed by calculation takes time). FIG. 20 is a flow chart showing an example of signal processing by Fourier transform using Jacobi operation, but the above signal processing is not limited to this.
In phase matching, it is often the case that translational phase matching is first performed on signals expressed in each orthogonal coordinate system, and then rotational phase matching is performed, but this is not the only option. They may be performed alternately or in the reverse order. Further, as described in Non-Patent Documents 13 and 15, each phase matching may be repeatedly performed (repetition is terminated when the corrected displacement amount becomes smaller than a preset threshold value). In some cases, the process of expansion and contraction is mixed.
Note that the local area and search area are not necessarily rectangular, and may have other shapes such as circles (data is stored in a rectangular array, such as a square, rectangle, cube, or rectangular parallelepiped array). In some cases, arrays of positions that fall outside the real-size circle, sphere, etc. may be zero-padded). The search area must be appropriately larger than the local area so that the signal circulated within the search area due to phase rotation by multiplying the complex exponential function in the frequency domain does not appear in the area of interest. It only increases the amount, and is appropriately determined a priori from the magnitude of the displacement vector of the observed object). Incidentally, when there is no displacement in the radial direction and the object to be observed is only displacement in the rotational direction, the size of the search area may be the same as that of the local area. Also, the search area may be provided in the previous frame. Moreover, in order to shorten the processing time, phase matching may be performed by discrete shifting of a digital signal instead of phase matching by multiplying a complex exponential function. In this case, it is not always necessary to apply motion compensation to the search area, and phase matching may be performed by directly searching (block matching) the local area within the search area of another frame.
In addition to the above-described phase rotation processing in the frequency domain of the polar coordinate system, the phase matching includes the DAS processing described in paragraph 0384 (the interpolation processing of method D1 or the processing of phase-rotating the analytic signal in the spatio-temporal domain of method D2. As in the case of shifting a multidimensional signal based on the delay processing of the shifting through the polar coordinate system, the multidimensional analytic signal represented by the polar coordinate system is processed and phase matching (radial shifting and rotation processing). The above phase rotation processing in the frequency domain of the polar coordinate system is based on the delay processing of method D3. These processes can also be used for DAS processing in the polar coordinate system, and can be performed during other beamforming such as Fourier beamforming, adaptive beamforming, minimum variance, etc. in the polar coordinate system, and the phase aberration correction of the signals described herein , motion compensation, phase matching (including novel processing methods), alignment, position correction, etc., for radial shifting and rotation in temporal, spatial, and spatio-temporal coordinates.
For example, in the two-dimensional case,
f(r+ _r0 , θ+ _θ0 )=f(r,θ)exp[j{kr( _r , _θ ) _r0 +kθ(r,θ)θ0 _} ] (2DASr)
In the case of three dimensions,
f(r+r ₀ , θ+θ ₀ , φ+φ ₀ )=f(r, θ, φ)
×exp[j{kr( _r , _θ ,φ) _r0 +kθ(r,θ, _φ )θ0+ _kφ (r,θ,φ)φ0 _} ] (3DASr)
can be calculated as However, (k _r , k _θ , k _φ ) are wave numbers in each direction at each position (r, θ, φ). In other words, in a digital signal, analog shifting can be applied in a discrete polar coordinate system (preferably with a high sampling frequency if high accuracy is required) using the wavenumber in each direction at each position. As in the one-dimensional case, it can be more accurate than method D1 (which corresponds to block matching or fragment matching of signals in the spatio-temporal domain), but less accurate than other methods (methods D3 and D4). but the calculation speed is fast. Suitable for applications requiring high computational speed.
It should be noted that these processes shift the point spread function as well as the observation target itself in the radial direction and the rotational direction at the same time (which can be a source of error in phase matching). These processes may also be performed on signals that do not have a carrier frequency (eg, image processing, etc.).

A Wiener filter can also be applied as Wp. In the spatio-temporal domain, the signal itself is directly weighted before imaging or displacement measurement. The signal is the signal r(x,y,z) before or after detection.

It should be noted that the noise signal n(x,y,z) can be statistically estimated by repeatedly collecting echoes of an object or calibration phantom. For example, variation can be used, and estimation may be performed locally by averaging or by ensemble averaging assuming a stationary process. In some cases, a typical value is a priori determined based on the subject, its state, experience of measurement, and the like. Moreover, it is not limited to this. In order to detect the signal r(x, y, z) for imaging, the equations (A12) and (A13) are multiplied at each position when performing envelope detection, square-law detection, and absolute value detection. Sometimes. Also, weighting can be performed in the same way when the analytic signal is multiplied by the conjugate of the analytic signal to obtain the power spectrum and obtain the self function. In addition, it can also be used as a signal preprocessing before using the autocorrelation method and Doppler method, which are displacement measurement methods using analytic signals, as well as the cross spectrum phase gradient method and cross-correlation method (in some cases, analytic signals are not used). Available.

When the signal is two-dimensional or one-dimensional, instead of r(x, y, z) and n(x, y, z) in equations (A12) and (A13), respectively, r(x, y) and n(x,y), r(x) and n(x). Also, in the region of interest obtained by scanning with each beam or each beam, equations (A12) and (A13) may be obtained and used globally. Instead of equations (A12) and (A13), equations (A4) to (A6) may also be used directly for echo weighting.

但し、ＰＷpn(ωx,ωy,ωz)とＰＷps(ωx,ωy,ωz)は、各々、ノイズと信号のパワースペクトラムであり、ＰＷps(ωx,ωy,ωz)には代わりにクロススペクトルの大きさの二乗(||Ｈp(ωx,ωy,ωz)||²）が使用されることがある。ｑは任意の正値である。
と表される重み付けを行って最小二乗化する。例えば、特許文献６の式（１）～（１４'）には、重み付けにクロススペクトルの大きさの二乗(||Ｈp(ωx,ωy,ωz)||²）そのものが使用された場合の記載があるが、その重みの代わりにＷp(ωx,ωy,ωz)が使用されることがあるわけである（別にビーム方向の変位のばらつきやＺＺＬＢが使用されることが有ることは上記の通りである）。尚、最小二乗化は、ｐ＝１～ＮのＮ個の波動又はビームの各々に関して評価されて、各関心位置において一度に最小二乗化される。 In particular, when using the multidimensional cross-spectral phase gradient method (Patent Document 6, Non-Patent Document 15), the Wiener filter can be applied not only in the spatio-temporal domain but also in the frequency domain. As described above, for each wave or beam, the frequency domain (ωx , ωy, ωz) using the method of least squares,

where PWpn(ωx,ωy,ωz) and PWps(ωx,ωy,ωz) are the power spectrums of the noise and signal, respectively, and PWps(ωx,ωy,ωz) is instead the magnitude of the cross spectrum The square (||Hp(ωx,ωy,ωz)|| ² ) may be used. q is any positive value.
The least-squares are obtained by weighting expressed as . For example, the equations (1) to (14′) of Patent Document 6 describe the case where the square of the magnitude of the cross spectrum (||Hp(ωx, ωy, ωz)|| ² ) itself is used for weighting. However, Wp(ωx, ωy, ωz) may be used instead of the weight (as mentioned above, the variation in displacement in the beam direction and ZZLB may also be used. be). Note that the least-squares is evaluated for each of N waves or beams, p=1 to N, and least-squared at each location of interest once.

The noise power spectrum PWpn(.omega.x, .omega.y, .omega.z) can be statistically estimated by repeatedly collecting echoes from an object or calibration phantom. In some cases, a typical value is a priori determined based on the subject, its state, experience of measurement, and the like. Moreover, it is not limited to this.

In addition, n(x, y, z)/r(x, y, z) represented in formulas (A12) and (A13), and PWpn(ωx,ωy,ωz)/PWps(ωx,ωy,ωz) is the reciprocal of the echo SN ratio SNRe, or the combined SN of SNRe and the correlation SN ratio SNRρ It may also be set based on the reciprocal of the ratio. Further, in a state with spatial resolution, in a region of interest obtained by scanning with each beam or with each beam, equations (A12') and (A13') are obtained globally, and equations (A12) and Similar to Eqs. (A13), Eqs. (A4)-(A6), it may be used directly for echo weighting (imaging or displacement measurement). Also, when detecting the signal r(x, y, z) for imaging (envelope detection, square-law detection, absolute value detection, etc.), formula (A12) and formula (A13) may be used. However, in that case, the square norm of the first spectrum Hp(ωx, ωy, ωz) (where the spectrum of the local signal or the signal covering the region of interest) in the equation need not be used. The same applies to obtaining an autocorrelation function signal by multiplying the spectrum Hp(ωx, ωy, ωz) by the conjugate of the local spectrum Hp(ωx, ωy, ωz) to obtain the power spectrum.

If the unknown displacement is a two-dimensional vector u = (Ux, Uy) or the unknown displacement is U (one) in the beam direction, H(ωx, ωy, ωz ), we can also use weights that are similarly expressed using respective Wiener filters on the cross-spectrum H(ωx, ωy) and H(ωx) between the signals before and after deformation or displacement.

Furthermore, when regularization is performed, the regularization parameter may be set using the above-described variations and the like in accordance with equations (10) and (10′).

When using the cross spectrum phase gradient method or other block matching method, it is possible to obtain displacement vectors in two or more directions using one wave or beam alone, and one wave or beam can be used alone. However, over-determined systems can still be constructed.

Moreover, when performing any of the above displacement measurements, it is also possible to perform measurement without using an over-determined system, and even in that case, the above weighting and regularization may be performed.

Displacement can be obtained from at least two signals, but when the displacement is large in the above displacement measurement (displacement of the point of interest or the local region containing the point of interest), the instantaneous phase (multidimensional or One-dimensional autocorrelation method, multi-dimensional or one-dimensional Doppler method, Non-Patent Document 13) and local phase (multi-dimensional or one-dimensional cross-spectral phase gradient method, Non-Patent Document 15) may be inverted. As already explained, phase matching (spatial shifting or phase rotation by multiplying a complex exponential function) may be performed instead of unwrapping the phase of the It is also an epoch-making treatment method that makes it possible, for example, C. Sumi et al, World Congress of Ultrasound in Medicine and Biology, Sapporo, 1994). In paragraph 0405, the shifting process in the radial direction and the rotational direction in the polar coordinate system by the same process has also been described. Also, various shifting processes that can be used for delay processing of DAS processing described in paragraphs 0384 and 0405 are also effective. Using cross-correlation-based multidimensional or one-dimensional cross-correlation methods (block matching is effective in this case) or cross-spectral phase-gradient methods (processing with coarse sampling intervals) that allow estimation of large displacements Coarse estimation is performed for phase matching (sometimes iteratively processed), and then fine estimation is performed using those methods (similarly iteratively processed, cross In the spectral phase gradient method, the sampling interval may be returned to the original value). As already explained, it may also be used for phase aberration correction.
Various processing can be performed by modifying these as a base (for example, the demodulation method described in Patent Document 7, etc.). Basically, the autocorrelation method uses the phase of the complex autocorrelation function, and to stabilize the estimate, it is obtained by multiplying the imaginary part/real part of the complex autocorrelation function by the inverse of the tangent according to Euler's formula. Either subject the instantaneous phase to moving average processing in time or space (Method Ai), or perform the same moving average on the complex autofunction and then multiply the imaginary part/real part by the inverse function of the tangent to obtain the phase. (Method Aii). In the case of the Doppler method, the instantaneous phase difference of the analytic signal itself is used, and in order to stabilize the estimation, the instantaneous phase obtained by multiplying the imaginary part/real part of the analytic signal by the inverse function of the tangent, or the difference thereof, Similarly, moving average processing can be performed (Method D). The cross-spectral phase gradient method uses the phase of the cross-spectrum of the local signal (Method C).
Among these, when coarse displacement is estimated using the cross-correlation method and phase matching (spatial shifting) is performed, when processing based on method Ai and method D, spatially inconsistent Since the instantaneous phase, which may have a continuous distribution, is subjected to moving average in time or space, the displacement cannot be obtained correctly (error occurs, and the final result of phase matching is a discontinuous displacement distribution. turn into). Therefore, after phase matching (spatial shifting), for example, in three-dimensional observation, the instantaneous frequency is (fx, fy, fz), and the unknown displacement to be updated with respect to the coarse estimation result (dx0, dy0, dz0) (ux, uy, uz), and the instantaneous phase before moving average processing, which may have a spatially discontinuous distribution, is θ. θ'=θ+fx dx0+fy dy0+fz dz0 is calculated using coarse estimation results (dx0, dy0, dz0) for the phase θ, and coarse estimation results (dx0, dy0, dz0) phase matching (corresponding to phase rotation) and moving average processing to obtain θ'', then solve fx dx+fy dy+fz dz=θ'' It is possible to obtain the estimation result of the unknown displacement (dx, dy, dz) directly (this is a new phase matching). In those calculations, the instantaneous frequencies (fx, fy, fz) may or may not be moving averaged. By this phase matching, it is possible to avoid adding incorrect estimated values of (ux, uy, uz) to coarse estimation results (dx0, dy0, dz0). It is also possible to solve the equation using θ' instead of θ'' without moving average of the instantaneous phase θ' after phase matching to obtain the estimation result of the unknown displacement (dx, dy, dz) directly. Yes, the result is the same as the result of adding the estimated value of the displacement amount (ux, uy, uz) to be updated to the coarse estimation result (dx0, dy0, dz0). The same applies to two-dimensional observation and one-dimensional observation.
On the other hand, when similarly coarse displacement is estimated using the cross-correlation method and phase matching (spatial shifting) is performed, when processing based on method Aii, a spatially discontinuous distribution There is no problem even if the moving averaged instantaneous phase has a spatially discontinuous distribution. . That is, for example, in the case of three-dimensional observation, when processing based on method Aii, a coarse estimation result is obtained with respect to the moving averaged instantaneous phase θ'' obtained after phase matching (spatial shifting). Phase matching that calculates θ'''=θ''+fx dx0+fy dy0+fz dz0 using (dx0,dy0,dz0) and adds phase for coarse estimation result (dx0,dy0,dz0) (corresponding to phase rotation) to obtain θ''', by solving fx dx+fy dy+fz dz=θ''', the estimation result of the unknown displacement (dx, dy, dz) is obtained directly (This is a new phase matching), and the displacement to be updated (ux , uy, uz) and add it to the coarse estimation result (dx0, dy0, dz0) to obtain the estimation result. In those calculations, the frequencies (fx, fy, fz) may or may not be moving averaged. The same applies to two-dimensional observation and one-dimensional observation.
In the case of processing based on method D, the cross-spectrum phase characteristics (frequency characteristic) may have a spatially discontinuous distribution, but there is no moving average applied to that distribution, so there is no problem (in each local region, the phase frequency characteristic becomes discontinuous there is no). Displacement equations may typically be set up simultaneously at the centroid frequency or frequencies in the vicinity thereof (in the case of one-dimensional, two-dimensional, and three-dimensional equations, at least one, two, and three frequency), or over-determined within the signal band. Even in this case, after phase matching (spatial shifting), for example, in three-dimensional observation, the frequencies in the signal band are (Fx, Fy, Fz), and the coarse estimation results (dx0, dy0, dz0) are If the unknown displacement to be updated is (ux, uy, uz) and the phase of the cross spectrum at the same frequency is α, the equation is expressed as Fx ux+Fy uy+Fz uz=α. On the other hand, using the coarse estimation result (dx0, dy0, dz0), calculate α'=α+Fx dx0+Fy dy0+Fz dz0 and add the phase for the coarse estimation result (dx0, dy0, dz0) By applying phase matching (corresponding to phase rotation) to obtain α', the estimation result of the unknown displacement (dx, dy, dz) can be obtained directly by solving Fx dx + Fy dy + Fz dz = α'. (new phase matching), instead of α' as before, solve the equation Fx ux+Fy uy+Fz uz=α using α to estimate the displacement (ux,uy,uz) to be updated and adding it to the coarse estimation result (dx0, dy0, dz0) also yields the estimation result. The same applies to two-dimensional observation and one-dimensional observation.

であり、尤度関数は、

従って、対数尤度Lは、式(LM３)に対数を掛けた

を得る。
ちなみに、超解像として、ボケ関数（線形時空間不変システム又は線形時空間変システムの点拡がり関数と考える）を表す行列Fと原像（ターゲット）を表すベクトルOとボケ画像を表すベクトルBとを用いてFO ＝ Bを解く場合に、最尤推定を行う場合には、同様に、

波動信号において直流を含まない場合には、式（LM9）の代わりに式（LM９'）の如く、平均値を零として計算することもある。ボケ画像の復元や超解像においては共分散行列の計算には時間を要することが多いので注意が必要である。
この時空間領域における処理は時空間変（spatially variant）な場合に有効である。物理開口面からの各距離の位置における点拡がり関数を同距離の位置において推定される複数の点拡がり関数（深さ方向又は横方向の1次元自己相関関数又は多次元自己相関関数）を平均化して推定することは有効である（非特許文献３５及び３６）。平均処理をせずに関心位置において推定された点拡がり関数を用いることもある。これらの時空間領域における処理には、例えば共役勾配法等を使用できる。この最尤推定を施した場合には、後述のＭＡＰ（Maximum a posteriori、例えば、非特許文献４２）が有効であるが、ＥＭアルゴリズムを用いた方法（非特許文献３１）も報告されている。これらの処理結果には、整形フィルタリングとして所望する点拡がり関数を時空間領域にてコンボリューションしたり、周波数領域においてその周波数応答を乗ずることも有効である。
時空間領域における処理の場合、点拡がり関数の時空間不変(spatially invariant)を仮定できる場合はデコボリューションを実施することになるが、計算速度を考えると、その処理は、本願に記載の周波数領域における逆フィルタリング処理を実施した方が良い。その場合に、点拡がり関数は時空間不変と考えられる領域における平均化されて用いられることがある。空間分解能がほぼ一様となる古典的な開口面合成や平面波送信を行って受信ダイナミックフォーカシングした場合等が該当するが、フォーカスビームを生成した場合の時空間変の場合でも、時空間不変を仮定して周波数領域において同処理を行うことがある。同様に、処理結果には、整形フィルタリングとして所望する点拡がり関数を時空間領域においてコンボリューションしたり、周波数領域においてその周波数応答を乗ずることも有効である。
また、a posteioriに変位の計測結果から分散を推定して正則化パラメータをそれに比例する様に決めたり（正則化、非特許文献１８）、その逆数を重み（方程式の信頼度）とする重み付き最小二乗法を実施する場合には、位相の上記エラーにより変位の計測結果がエラーを含んでいるとエラーを生じる。
さらに、それらの統計量は、coarseな推定結果(dx0,dy0,dz0)に対して更新すべき変位量(ux,uy,uz)に関する方程式に実施されることになる。従って、１方向の速度計測法である１次元自己相関法（非特許文献２０）にて導出されているばらつき（パワードプラにて使用されている）や、これを基に本願にて多次元自己相関法に関して導出されているばらつきや、Ziv-Zakai Lower Bound（ＺＺＬＢ）は、a prioriであり、変位計測に関して定常過程を仮定していないが、同じくエラーを生じる。この種の最適化処理を実施する場合は全ての変位計測法において位相の空間分布を連続にする上記の位相マッチングを施した上で処理する必要がある。位相マッチング中において更新すべき変位（残差変位）そのものやその分の位相データに移動平均処理や最適化を施してはいけない。また、この新しい位相マッチングは、その必要が無くとも、数種類の最適化法を用いて複数の推定結果を得たり、それらを統合又は判断して最終的な結果を得ることも有り、計算量が増えるがその様にしておくと良いことがある。
ここで、最尤推定において有効なMAP推定について記載し、さらに、正則化との類似性について述べる（非特許文献４２）。
式（LM１）におけるDのMAP推定は、Dのa posteriori確率密度関数

但し、p(θ|D)はθのa posteriori確率密度関数（最尤推定）の式（LM３）であり、

はDの確率密度関数である。
を最大化して実施できる。このＭＡＰ推定は、重み付き最小二乗法としてコスト関数

を最小化することと等価であり、解は、

但し、Pは正則化オペレーターであり、λはハイパー正則化パラメータである。
を解くこととなる。正則化オペレーターPとしては単位行列を用いることができ、この場合の正則化パラメータλは空間的に可変（variable）に使用することが可能である（非特許文献１７）。この正則化オペレーターは全帯域の通過型フィルタであるので、Dの結果は真値に比べて小さくなることがある。別の正則化オペレーターPとして、高域通過型フィルタや微分フィルタ(勾配作用素Gを用いたラプラシアン作用素G^TGやそのn乗ノルム)等が使用されることもあり、この場合は高域周波数のエラーが効果的に抑圧されるが、高周波成分は失われることがある（しかし、ボケ画像の復元や超解像にも使用できる）。複数の正則化オペレーターを同時に使用することもある。導出される式(REG１)の通りではなく、便宜上、λを正則化パラメータとして典型値を用いたり、新しい位相マッチング時のθの分散や式（LM１）の左辺と右辺の差の分散、Dの分散に比例する様に定める場合もある（正則化パラメータを大きくすると正則化の効果が増す）。また、式（A２）の重み行列Wを用いた重み付き最小二乗法は、W^TWにDの分散の逆数を用いたものである（Dのばらつきが小さく精度が高い式は重視して重みを大きくする）。
式（MAP２）と式(REG１)との類似性から、両式を混ぜて使用することも有効である。これらにおいて、E[D]が零ベクトルで無くとも、便宜上、零ベクトルとして計算することもある。
尚、ボケ画像の復元や超解像においては、上記のシステムFO＝Bにおいて最尤推定を施すと、式（LM７）~（LM９）を解くことと成り、また、上記の変位計測と同様にMAP推定を行ったり、ＥＭアルゴリズムを用いることも可能であり、また、時空間不変システムとして周波数領域において処理することも有効である。上記の変位計測の場合と同様に時空間領域のシステムを正則化して安定化して解くことも有効である（非特許文献４３）が、同文献にて公開されている周波数領域における処理によって高速処理が可能である。

但し、＊は共役を表し、ここでは各行列やベクトルで表される空間分布の周波数応答を同一の文字を用いて表している。同文献には、通常のウィーナーフィルタよりも正則化が有効である結果が報告されいる。本発明においては本願に記載のウィーナーフィルタの応用（変則型を含む）や時空間的にvariantな正則化（正則化パラメータが時空間的にvariant）が行われる。また、上記の如くに整形フィルタが施されることがある。
この正則化は、受信信号のＳＮ比が低い場合や、超音波パラメータやビームフォーミングパラメータが異なる信号間や、物理的に異なるセンシング信号間の処理等においては、空間分解能が低下することを代償としても、この様な平滑化処理が有用となる可能性がある。
その他として、例えば、段落０３７７に示した逆解析や他の様々な逆解析（システムがAx＝bと表される場合）においても同様に最適化することが可能である。最尤推定やＭＡＰ、正則化以外にも、ベイズ推定を同様に実施できるし、その他にも様々な有効な方法を施すことができる。ちなみに、変位ベクトルや電流密度ベクトル等のベクトルが観測対象の場合は各方向成分に関して別の正則化パラメータや別の正則化オペレーターが使用されることがある（非特許文献１７）。 Multiple waves with different waves and beamforming parameters generated at the same position (in addition to physically generated waves, pseudo-waves generated by superimposing waves with different waves and beamforming parameters, (including pseudo-waves generated by frequency-dividing the spectrum) are derived one by one, and the same number of equations as the unknown displacement components are simultaneous, or the same number or more of the equations are simultaneously over -Determined systems may be realized, or equations derived at different positions may be solved simultaneously. may be processed.
In particular, when performing the optimization processing described in this application (paragraph 0402, paragraph 0403, etc.) or other optimization processing in order to increase the accuracy or stabilize the displacement measurement (estimation), all displacement measurement methods should be processed after applying the above-described phase matching that makes the spatial distribution of the phase continuous. First, assuming a local stationary process for the phase of the complex autocorrelation function, the instantaneous phase difference of the analytic signal, and the phase of the local cross spectrum, estimate the temporal or spatial local mean, variance, or covariance. When used with , an error will occur if the phase contains the above error. This is the case, for example, when performing maximum likelihood estimation (sometimes implemented with Maximum a posteriori (MAP)). A normal system of equations or an over-determined system of equations at the point of interest is

and the likelihood function is

Therefore, the log-likelihood L is obtained by multiplying equation (LM3) by the logarithm

get
By the way, as super-resolution, the matrix F representing the blur function (considered as a linear spatio-temporal invariant system or the point spread function of a linear spatio-temporal varying system), the vector O representing the original image (target), and the vector B representing the blurred image Similarly, when solving FO = B using

If the wave signal does not contain direct current, the mean value may be calculated as zero, as in formula (LM9') instead of formula (LM9). It should be noted that it often takes time to calculate the covariance matrix in restoration of blurred images and super-resolution.
Processing in this spatio-temporal domain is effective in spatially variant cases. The point spread function at each distance from the physical aperture plane is averaged from multiple point spread functions (one-dimensional autocorrelation function or multi-dimensional autocorrelation function in the depth direction or lateral direction) estimated at the same distance position. It is effective to estimate by A point spread function estimated at the location of interest without averaging may also be used. A conjugate gradient method, for example, can be used for processing in these spatio-temporal domains. When this maximum likelihood estimation is performed, MAP (Maximum a posteriori, for example, Non-Patent Document 42) described later is effective, but a method using the EM algorithm (Non-Patent Document 31) has also been reported. For these processing results, it is also effective to convolve a desired point spread function as shaping filtering in the spatio-temporal domain, or to multiply it by its frequency response in the frequency domain.
For processing in the spatio-temporal domain, if one can assume spatially invariant of the point spread function, then decovolution will be performed, but given the computational speed, the processing can be performed in the frequency domain as described herein. It is better to perform the inverse filtering process in . In that case, the point spread function may be used after averaging over a region that is considered to be spatio-temporal invariant. This applies to the classical aperture plane synthesis where the spatial resolution is almost uniform, and the case of receiving dynamic focusing by performing plane wave transmission. The same processing may be performed in the frequency domain. Similarly, it is effective to convolve the processing result with a desired point spread function as shaping filtering in the spatio-temporal domain, or to multiply it by its frequency response in the frequency domain.
In addition, the variance is estimated from the displacement measurement results a posteiori and the regularization parameter is determined to be proportional to it (regularization, Non-Patent Document 18), or the reciprocal is weighted (reliability of the equation). When performing the least-squares method, an error occurs if the displacement measurement result contains an error due to the phase error.
Further, those statistics will be implemented into the equations for the displacements (ux, uy, uz) to be updated for the coarse estimation results (dx0, dy0, dz0). Therefore, based on the variation (used in power Doppler) derived by the one-dimensional autocorrelation method (Non-Patent Document 20), which is a unidirectional velocity measurement method, the multidimensional self The scatter derived for the correlation method, or the Ziv-Zakai Lower Bound (ZZLB), is a priori and does not assume a stationary process for the displacement measurement, but it also gives rise to errors. When performing this type of optimization processing, it is necessary to perform the above-described phase matching that makes the spatial distribution of the phase continuous in all displacement measurement methods. During phase matching, the displacement (residual displacement) itself to be updated and the corresponding phase data should not be subjected to moving average processing or optimization. In addition, even if this new phase matching is not necessary, it is possible to obtain multiple estimation results using several types of optimization methods, and to obtain the final result by integrating or judging them, so the amount of calculation is large. Although it will increase, it is good to keep it that way.
Here, we describe MAP estimation, which is effective in maximum likelihood estimation, and describe its similarity to regularization (Non-Patent Document 42).
The MAP estimation of D in equation (LM1) is a posteriori probability density function of D

However, p(θ|D) is the formula (LM3) of the a posteriori probability density function (maximum likelihood estimation) of θ,

is the probability density function of D.
can be implemented by maximizing This MAP estimation is performed as a weighted least squares method with the cost function

is equivalent to minimizing , and the solution is

where P is the regularization operator and λ is the hyper-regularization parameter.
will be solved. A unit matrix can be used as the regularization operator P, and the regularization parameter λ in this case can be spatially variable (Non-Patent Document 17). Since this regularization operator is a full-pass filter, the result of D can be small compared to the true value. As another regularization operator P, a high-pass filter or a differential filter (Laplacian operator G ^T G using the gradient operator G or its n-th power norm) may be used. Errors are effectively suppressed, but high frequency components can be lost (but it can also be used for blurry image restoration and super-resolution). Multiple regularization operators may be used simultaneously. Instead of the derived formula (REG1), for convenience, we use a typical value for λ as a regularization parameter, the variance of θ at the time of new phase matching, the variance of the difference between the left side and the right side of formula (LM1), and the value of D In some cases, it is determined to be proportional to the variance (increasing the regularization parameter increases the regularization effect). In addition, the weighted least squares method using the weight matrix W of formula (A2) uses the reciprocal of the variance of D for W ^T W. ).
Due to the similarity between formula (MAP2) and formula (REG1), it is also effective to use a mixture of both formulas. In these, even if E[D] is not a zero vector, it may be calculated as a zero vector for convenience.
In addition, in the restoration of blurred images and super-resolution, if the maximum likelihood estimation is performed in the above system FO = B, the equations (LM7) to (LM9) can be solved, and similarly to the above displacement measurement It is also possible to perform MAP estimation or use the EM algorithm, and it is also effective to process in the frequency domain as a spatio-temporal invariant system. As in the case of displacement measurement above, it is also effective to regularize and stabilize the system in the spatio-temporal domain (Non-Patent Document 43). is possible.

However, * represents a conjugate, and here, the frequency response of the spatial distribution represented by each matrix or vector is represented using the same character. The same document reports the result that regularization is more effective than the usual Wiener filter. In the present invention, the application of the Wiener filter described in the present application (including an irregular type) and spatio-temporal variant regularization (the regularization parameter is spatio-temporally variant) are performed. Also, a shaping filter may be applied as described above.
This regularization comes at the cost of lower spatial resolution when the signal-to-noise ratio of the received signal is low, between signals with different ultrasound parameters or beamforming parameters, or between physically different sensing signals. may also benefit from such smoothing.
In addition, for example, the inverse analysis shown in paragraph 0377 and various other inverse analyzes (where the system is expressed as Ax=b) can be similarly optimized. In addition to maximum likelihood estimation, MAP, and regularization, Bayesian estimation can be performed similarly, and various other effective methods can be applied. Incidentally, when a vector such as a displacement vector or a current density vector is to be observed, another regularization parameter or another regularization operator may be used for each directional component (Non-Patent Document 17).

This new phase-matching process can generate different waves at the same position and multiple waves with beamforming parameters (including physically generated waves and pseudo-waves generated by frequency-dividing the spectrum). ), pseudo waves generated by superimposing those waves (broadband and high resolution by coherent addition), or phase aberration correction during beamforming processing (high precision by applying displacement measurement method It is also effective in high-precision phase matching based on phase aberration measurement), and signal separation by ICA (Independent Component Analysis) processing (which has a more effective multiplexing effect than averaging) and super-resolution by nonlinear processing etc. is also effective. When there is tissue displacement in their phase aberration correction, motion compensation can also be performed at the same time. It is also effective to perform phase aberration correction during or after beamforming, and to perform super-resolution during or after beamforming for received signals in multiple frames, which are often used in multiple frames. processing such as applying an image, superimposing the results of super-resolution, etc.). It is also effective to process these between different sensing signals such as MRI, ultrasound, X-ray CT, OCT, and terahertz. In such processing using multiple frame data, there is a lack of signal in large displacements and deformations of the observed object, especially in uncontrolled spontaneous displacement of the observed object (for example, when the observed object is one-dimensional, two-dimensional, or three-dimensional). Because the accuracy of phase matching between multiple frame data decreases due to physical effects such as heating (treatment) and chemical effects such as chemotherapy, the displacement measurement described in this application (shifting optimization such as regularization and maximum likelihood estimation are effective, but the new phase matching is also useful in these cases. Sometimes speckle signals are processed, sometimes specular signals are processed, and edge extraction, enhancement, and feature points (for example, tissue structures such as blood vessel bifurcation positions when observing humans) are processed as necessary. Displacement tracking, etc., which are obtained and performed, are effective. Alternatively, depending on the signal, it may be effective to positively smooth the signal in the above regularization, or perform processing after performing envelope detection.
In this way, the new phase matching processing includes phase matching in displacement measurement, etc., DAS processing (methods D1 to D3) described in paragraph 0384, and other methods such as Fourier beam forming, adaptive beam forming, and minimum variance. Delay estimation (phase aberration correction) in beamforming, phase aberration correction, motion compensation, alignment, position correction, etc. during beamforming described in paragraph 0372 Signal shifting in time coordinates, space coordinates, and spatiotemporal coordinates Useful for estimating quantities.

There are also various techniques for detecting and imaging the movement of an object based on the observed wave motion. , speed information, presence or absence of motion, complexity, etc. In some cases, a contrast medium (microbubbles) is actively used to enhance the intensity of waves from blood in blood vessels and heart chambers, and measurement imaging is performed. It may be effective not only for morphological observation but also for functional measurement. A typical example of a self-emanating contrast agent is a radioisotope used in PET (Positron Emission Tomography), and observation is performed based on counting the number of occurrences. This is the type targeted by the passive device of the second embodiment. Vibration can also be applied to generate a magnetic field. Therefore, it is mechanically stimulated by the transmitting transducer, which is the transmitting means, and the electromagnetic waves in response are observed by the receiving transducer, which is the receiving means. Photoacoustic, etc., as described above, are also possible. For example, ¹⁸ FDG ( ¹⁸ F-fluorodeoxyglucose, glucose labeled with a positron-emitting nuclide), a representative PET contrast agent, shows that cancer cells take up more glucose (3 to 20 times) than normal cells. It is used for the early detection of cancer in the whole body by applying its properties, but by applying photoacoustics to this, it is possible to receive ultrasonic waves and detect cancer at an early stage (the accumulation mechanism by metabolic trapping is Like glucose, it is taken up into the cell membrane via the glucose transporter in the cell membrane and metabolized by the enzyme hexokinase, but unlike glucose, it remains in the cell without proceeding to glycolysis). Sometimes it is used in conjunction with PET, but sometimes ultrasound is done separately from PET. The processing itself is based on receive beamforming for incoming ultrasonic waves (transmitted waves) using triggers based on the timing of laser irradiation. A positron is emitted from a positron-emitting nuclide (positron nuclide) by β? Although the spatial resolution is low due to the observation of gamma rays (ie, annihilation radiation), there is the advantage that the generated ultrasonic waves can be observed at a relatively high resolution. It is also suitable for early detection, and it is possible to observe the presence or absence of disease based on the presence or absence of accumulation, and to judge the degree of malignancy and progression by observing the degree of disease from the intensity of the photoacoustic signal ( Glucose metabolism is accelerated in malignant tumor cells, resulting in high accumulation). In addition to the D-type glucose present in the body, "Applying the characteristic that cancer lesions specifically take up L-type glucose to fluorescent L-glucose", the Hirosaki University Graduate School of Medicine Integrative Functional Physiology Course (former Physiology Department) outside the field of photoultrasonics One lecture) The use of L-glucose is also effective, following the results of Katsuya Yamada's research group (oral or injection intake). In addition, for example, the brain is the most active in sugar metabolism in the human body, and Alzheimer's dementia is accompanied by metabolic abnormalities from an early stage. In addition, when ischemia progresses, myocardial glucose metabolism is enhanced (positive), but when necrosis occurs, glucose metabolism does not occur. Photoacoustic can also be used for those observations. In the case of the brain, the craniotomy may be performed, and when targeting the disease of the deep organs of the abdomen, laparotomy, laparoscope (laparoscopy), catheter, etc. may be used in the vicinity of the disease, such as laser irradiation and Ultrasound detection may also be performed. In many of these applications, the light absorption frequency characteristics of the contrast agent have been clarified and are publicly known. , Observe the ultrasonic waves generated during broadband or specific band laser irradiation in a wide band as follows, or focus on a specific ultrasonic frequency band, or It is possible to observe sound waves in a wide band, focus on a specific ultrasonic frequency band at that time, and simulate broadband laser irradiation by superimposing the observed ultrasonic signals. The ultrasonic sensors may be arranged in a one-dimensional, two-dimensional, or three-dimensional array with the same broadband or specific band characteristics, or with different band characteristics in a similar array. They may be arranged in a pattern (for example, locally continuously, alternately, or periodically), and the light source and ultrasonic sensor may have separate bodies or may be integrated. (Sometimes the light source and ultrasonic sensor are replaced and they are installed in the same place), and imaging is not always performed, and only quantitative observation based on numerical values is possible. It is sometimes done. As described above, the observation target is the glucose concentration, and it can be applied to the observation and imaging of the blood sugar level. In addition to sugars, PET contrast agents include oxygen, water, amino acids, nucleic acids, neurotransmitters labeled with positron-emitting nuclides ( ¹¹ C-methionine, ¹¹ C-acetic acid, ¹¹ C-choline, 11 C-choline, ¹¹ C-methylspiperone, ¹³ N-ammonia, ¹⁵ O-water, ¹⁵ O-oxygen gas, ¹⁸ F-fluorodova, etc.) have been put to practical use, and those developed for photoacoustics have been developed separately. It can also be used for various contrast agents. In these photoacoustics, Doppler measurements (including vector measurements) may be performed on blood, urine, and body fluids with and without contrast agents, and soft and hard tissue Displacement, deformation, viscoelasticity, and thermodynamic properties of organs, including , may be observed (including not only diagnosis and examination but also various treatments). When exposed to radiation, as usual, it is necessary to prevent exposure or reduce the exposure dose to a safe value or less. Contrast agents may also have therapeutic effects. Conversely, a therapeutic agent may have the effect of a contrast agent.
Various contrast agents may be used at the same time, but there are also cases where they are not used. Ultrasonic sensors with broadband reception characteristics are actively used, and broadband photoacoustics signals are output at once for the above applications, etc. , or by filtering (analog or digital) to select a frequency band for the above applications (including imaging of the Photoacoustics signal itself). Observations of the above applications using Photoacoustics signals for each frequency or frequency band may be superimposed or averaged. In the case of imaging, a color may be assigned to the magnitude of the numerical value of the observation data to be displayed, or shading may be added even within the same color to provide resolution. In addition, observation results in each observed frequency or frequency band may be imaged and displayed in a superimposed manner. In that case, different colors may be assigned to the observation results of each observed frequency or frequency band, and resolution may be given to the numerical values of the observation data displayed with shading. It is important to understand which contrast agent or substance to be observed is responsible for the photoacoustics signal in each frequency band (it is also important to measure the dispersion of photoacoustic waves, but for example, human tissue or human diseased tissue , Absorbance data from 400 nm to 25 μm of related substances are abundant and useful), and photoacoustics signals in the frequency band that match those targets are sometimes actively used, but not necessarily. It is useful (that is, when multiple markers are mixed, such as at least one originally existing in the observation target, at least one taken in from outside the observation target, and when they exist together, especially irradiating light with a bandwidth in which a Photoacoustics signal is generated at them, such as when there are no markers without being aware of the markers). It may be imaged as described above, or may be used for observation of displacement and temperature described in this application, and other inverse analysis (observation) using them. Further, as described herein, spectral analysis (analog or digital filtering), signal processing such as ICA (with or without phase aberration correction), and other image processing (e.g., deterministic or stochastic In some cases, the signals from the different markers are separated and used by using a standard image pattern as an index. It is also effective to use the difference in signal strength or spectrum magnitude (rms value) itself. It can also be the analysis of the marker itself. In that case, the signal when a contrast agent is used and when it is not used may be targeted. Similarly, it may be imaged as described above, and the separated signals may be used for displacement and temperature observations, as described herein, and other inverse analyzes (observations) using them. In addition, for example, when imaging fluid in a medium that is stationary or moving slowly, such as blood flow, or observing their movement, the photoacoustics signal itself or ultrasonic waves (echoes) that may be used in combination Using Doppler observation on the signal, it is possible to separate directly using the clutter rejection filter and signal intensity difference used in normal blood flow Doppler observation, or to separate by specifying the blood flow position (area). is also useful. Of course, this processing is effective even when observing by appropriately setting the frequency and band of the irradiation light to the target fluid. It is also useful to use Doppler observation using an optical device (optical pulse or electromagnetic wave pulse) for processing. Importantly, it also allows a detailed evaluation of the medium surrounding the separated fluid (displacement, temperature, inverse analysis, etc.). These are useful methods for continuous analysis/analysis and synthesis (theory).
For these photoacoustics signals (received signals), well-known beamforming methods as well as various beamforming methods and signal processing described herein are used. Applications therefore include imaging of the Photoacoustics signal itself, as described above. For detection processing, the method described in this specification is used in addition to the known method. Various signal processing techniques can be used to separate the signals, such as ICA as described herein, but for example, the spectrum can be split, split after beamforming, or angular spectrum before beamforming. may be divided. When dividing the angular spectrum, in addition to Fourier beamforming, various beamforming such as DAS processing can be performed. In superposition, coherent addition or incoherent addition is performed. In particular, by using the beamforming method and signal processing described in this specification, weak signals can be processed without leakage, and detailed observation can be performed with high accuracy. Photoacoustics signals are basically broadband, so it is possible not only to realize high-resolution observation (including the above applications), but also to apply it as a microscope. It can also be realized. In this case as well, the above application and the like are performed. In addition, the light irradiation can be performed by irradiating a wide range at once, such as plane waves, spherical waves, or cylindrical waves, or by scanning the light beam (mechanical scanning or electronic scanning), and the latter is better than the former. High-resolution observation is possible.
In addition, when the reception characteristics of the ultrasonic sensors are narrowband, each photoacoustics signal received by using a plurality of ultrasonic sensors with different reception bands may be used as described above, or may be used in combination. sometimes.
These observed photoacoustic signals are similarly processed instead of ultrasonic echoes, subjected to various signal processing (including inverse analysis) described herein, and variously applied. For example, Doppler observation, dynamic reconstruction, thermodynamic reconstruction based on temperature observation, etc. are not limited to these.

In addition, the wave is separated before the final inverse Fourier transform and detected (square detection, envelope detection, etc.), separated after the final inverse Fourier transform and detected simultaneously, or separated from the beginning can be detected before and after the Fourier transform (Non-Patent Document 1), and the intensity distribution of each wave can be visualized, or the incoherent signals obtained by the detection can be superimposed. , which emphasizes deterministic signals (e.g., reflected and specular signals) and reduces stochastic signals (e.g., scattered and speckle signals) to provide a spatial representation of structures in objects and media. (Past invention of the inventor of the present application).

In some cases, coherent signals with superimposed waves are detected and their intensity distributions are imaged. In addition, it is also possible to image the undetected coherent signal and image the waving itself, or to present the image of the phase distribution obtained together with the image of the intensity (amplitude) of the signal. A single wave may be similarly imaged.

Gray images and color images are the most popular display methods, but when quantification is required at that time, numerical values corresponding to the displayed brightness and color may be displayed as bars. be. In addition, it may be displayed as a bird's-eye view or the like, and may be displayed using CG together. They may be displayed as still images or videos, videos may be displayed frozen, they may be displayed in real-time, or they may be processed and displayed offline. There is also Wave data or image data stored in a storage device (or storage medium) may be read and displayed. Any numerical value over time may be displayed as a graph.

In addition, for example, it is possible to observe the temperature distribution of the measurement target using microwaves, infrared rays, or signals in the terahertz band. The waves transmitted to the radiating object are modulated and detected (even in the passive type apparatus according to the second embodiment, the temperature distribution of the target is measured from the radiated waves themselves. sometimes called). Spatial resolution can be obtained by using pulse waves, burst waves, and beamforming, instead of continuous waves, as with other wave motions. Some people think that it is limited only to the surface of the object), but if microwaves and terahertz are used, the temperature distribution inside the object can also be observed. Based on these observed physical quantities and chemical quantities, high-order processing such as an inverse problem approach is performed to obtain viscoelastic modulus, elastic modulus, viscosity (Patent Document 9), , electrical properties (conductivity and permittivity, Patent Document 8), magnetic permeability, wave propagation speed (light speed, sound speed, etc.), attenuation, scattering (forward scattering, backscattering, etc.), transmission, reflection, refraction, surface waves, Alternatively, wave sources and the like may be sought including frequency dispersion. Patent Documents 8 to 10 disclose a method capable of reconstructing the distribution of physical properties related to one physical quantity (observed quantity) within a region of interest. In addition, the distribution of physical properties may be evaluated directly from the waves themselves that are directly sensed by sensors. In medical applications, including the use of ultrasound and MRI, observation of viscoelastic modulus in addition to temperature and thermophysical properties of cancerous lesions, inflamed areas during heating/heat treatment, after such treatment and surgery. and monitoring may be carried out. In addition, body temperature observation (morning, noon, night, metabolism, growth, aging, before and after eating, before and after smoking, load on the peripheral system, including electrophysiological nerve control, etc.), exercise load on various organs, etc. can be observed. It is not limited to medical applications, and other organic and inorganic substances, or mixtures of them, can be observed. may be implemented. The application of terahertz can be applied not only to these measurements, but also to imaging of transmitted waves, reflected waves, etc., as well as other wave motions, and can also be applied to Doppler measurements and the like. Characteristically, it can also be applied to observation of inorganic substances, like X-rays. Other waves can also be used to observe organic and inorganic matter, but they can also be used simultaneously with other waves for fusion.

Measured physical quantities such as displacement and temperature may also be displayed as images in a similar manner, and may also be displayed superimposed on the morphological image obtained at the same time. Images representing these distributions are often required to be quantitative, and numerical values corresponding to displayed brightness and color are sometimes displayed as bars. In addition, it may be displayed as a bird's-eye view or the like, or may be displayed using CG together. They may be displayed as still images or videos, videos may be displayed frozen, they may be displayed in real-time, or they may be processed and displayed offline. There is also Wave data or image data stored in a storage device (or storage medium) may be read and displayed. Any numerical value over time may be displayed as a graph.

In addition, additional information on waves to be observed may be provided from other devices through input devices, and observation data of other physical and chemical quantities may be provided. In addition to processing, data mining, independent signal separation (independent component analysis), principal component analysis, code, multidimensional spectrum analysis, signal separation by MIMO, SIMO, MUSIC, signal separation based on identification of objects by parametric methods, High-order processing such as super-resolution or ISAR (Inverse synthetic aperture), which may use these methods together, may be performed. In other words, super-resolution may be applied to beamformed (including aperture plane synthesis), and before reception beamforming or without beamforming at all (transmission and reception signals for aperture plane synthesis) Beamforming may be performed after performing super-resolution processing, and in each case, these methods may be used together.

These processes are also performed in the passive device according to the second embodiment, and will be described in detail there. is transmitted and scanned, it is greatly different that the position of interest can be specified in the received signal. If there is, we can understand them by modulating and demodulating waves with information about the wave source with high resolution, and also plane waves (flat array), cylindrical waves (ring-shaped array), spherical waves (spherical shell-shaped array) If you use a wave similar to , it is possible to capture them at high frame rates and at high speed.

When performing communication, it is possible to narrow down the position of the communication destination from the transmission transducer, thereby improving energy saving and communication security. When performing observation, there is a high degree of freedom in constructing the measurement system. In addition, in the processing based on those system theory, it is easy to identify the point spread function to be generated and to adjust it according to the purpose. It is also possible to use multiple transmitting transducers and receiving transducers (sometimes also serving as transmitting transducers), and the waves to be transmitted and received may be the same or different, and sometimes As such, a plurality of device main bodies dedicated to a plurality of transducers that operate in synchronization may be used, or a passive type device according to the second embodiment may be used in combination. may be connected to other devices, including a device that controls the , through a dedicated or normal network, and the device itself may have a network control function.

Based on various observation data, equipment such as equipment for manufacturing materials and structures, equipment for treatment and repair, equipment for application (robots, etc.) may operate in conjunction, but it is not limited to these. . Measurements and high-level calculations using these waves may be carried out by separate devices using wave data stored in removable storage devices (storage media). Data may also be stored in the same type of storage device (storage medium) and used in another device.

If the received signal received and stored in a memory or storage device (storage medium) contains harmonic components generated in the object or medium, the fundamental wave and the harmonic components are Waves (sometimes only the second harmonic, sometimes multiple harmonics without ignoring higher-order components) are separated and beamforming processing (normal phased addition) is sometimes performed. and may be separated after beamforming is performed. There is a separation method that separates the spectrum in the frequency domain, but the bands may overlap, and in the so-called pulse inversion method in medical ultrasound, waves with opposite polarities are generated in the same phase. In the device according to this embodiment, if the received signals for each transmission are superimposed before or after beamforming, the second harmonic component can be extracted and the fundamental wave can also be obtained.

In addition, a separation method using a polynomial is also known, and in the apparatus according to this embodiment, one-dimensional processing in the wave propagation direction, or when the wave is modulated in the lateral direction, and strictly speaking, the wave propagation direction is different Multi-dimensional processing can take into account changes in position and can be done before or after beamforming. However, in the case of performing beamforming after separation, when beamforming is performed for each of the fundamental wave and harmonics, each beamforming process is performed, which requires computation time. Processing may be performed, but separation after beamforming is faster.

On the other hand, when the wave transmitted from each transmission aperture is coded, when beamforming is performed, the wave generated by transmission from which transmission aperture element corresponds to the received signal received by each reception aperture element. It is widely known that the components are separated by signal detection based on matched filters to perform not only reception dynamic focusing but also transmission dynamic focusing. This is also effective in high-speed transmission by plane wave transmission, and is also effective in focusing beams and steering.

In addition, when transmitting a plurality of waves or beams at the same time, for example, waves having a plurality of different frequencies and a plurality of different propagation directions are coded and transmitted, and the received signal are similarly decoded to separate the received signals produced by their respective waves or transmit beams, thereby improving resolution. This is effective, for example, when the propagation directions are the same due to waves with overlapping bands, refraction, reflection, transmission, scattering, or the like. Based on the basic idea of encoding the separating waves with a different code. It may simply be coded under the same physical parameters.

Although it is possible to solve simultaneous equations in these methods, the processing speed is high and the effect of the matched filter is also obtained. Codes have also been developed that are suitable for the subject and medium. However, as the number of elements used increases, the code length becomes longer and the energy of the signal becomes larger, and it is important to make effective use of this. is known to be unsuitable due to a decrease in , and similar problems arise in chirp signal compression.

In communication, waves transmitted from each transmission aperture element are coded with information encoded and transmitted (for example, plane waves, cylindrical waves, or spherical waves are used for beamforming). Focusing, which may be in multiple locations, to ensure accuracy at those locations, and may also ensure security or save energy in order to communicate with local or specific targets. ), beamforming is performed on the received signal, and decoding is performed. Applications of encoding in the apparatus according to the present embodiment are not limited to these, but these processes are performed in a digital signal processing unit (some include memory-embedded units), memory, and storage devices (storage media). .

In addition, physical quantities (displacement, velocity, acceleration, strain, strain rate and other magnitudes and directions, temperature, etc.), chemical quantities, or additional information observed by the device according to the present embodiment or provided by others , In addition, the viscoelastic modulus, elastic modulus, viscosity, thermophysical properties, electrical properties (conductivity Permeability, wave propagation velocity (light speed, sound speed, etc.), attenuation, scattering, transmission, reflection, refraction, surface wave, wave source, material, structure, or their frequency dispersion, etc. beamforming parameters (transmission intensity, transmission and reception apodization, transmission and reception delay, steering angle, time interval between each transmission and reception (scan rate), frame rate) , the number of scan lines, the number, shape, size, orientation and position of the effective aperture, the shape, size and orientation of the aperture element, the orientation of the physical aperture, or the polarization mode, etc.) may be optimized. Thereby, for example, certain targets can be detected (based on shape, material, structure, motion characteristics, temperature, humidity, etc.) with uniform spatial quality (spatial resolution, contrast, scan rate, etc.) Scattered waves (forward or backscattered waves), transmitted waves, reflected waves, depending on the movement, composition and structure of the object, with high quality (spatial resolution, contrast, scan rate, etc.) at the detected and related positions Optimized beamforming is performed, such as properly capturing waves, refracted waves, or surface waves, and focusing on wide-angle observations.

The wave propagation velocity is determined by the physical properties of the medium, but the physical properties change depending on the environmental conditions such as pressure, temperature, and humidity. is. The propagation velocity may be measured in real time, or the propagation velocity may be determined from observed environmental conditions based on calibration data for environmental conditions. The apparatus according to this embodiment further includes a phase aberration correction unit that corrects the non-uniformity of the propagation velocity. Phase aberration correction can be performed. Also, after reception, correction by complex exponential multiplication in the frequency domain is possible in the digital signal unit to correct for propagation velocity inhomogeneities in the transmission and/or reception propagation paths. Alternatively, correction can be applied directly in the calculation of the Fourier transform and the inverse Fourier transform. The reliability of the measured propagation velocity is determined by generating an imaging signal for the target object itself or a reference object that exists or is installed near the target object, and evaluates the state of imaging, spatial resolution, signal strength, contrast, etc. It can be confirmed as an index, and may be further adjusted based on this. Also in the passive second embodiment described later, transmission and/or reception phase aberration correction may be performed after reception.

A wave propagates and expands while being affected by attenuation, scattering, transmission, reflection, refraction, etc. Basically, the intensity of the wave weakens with the propagation distance. Therefore, the apparatus according to the present embodiment is equipped with a function of correcting the attenuation of the signal before or after beamforming based on Lambert's law, or the operator can input attenuation from the input device. A function may be provided to adjust the correction at each position or distance, and as described above, a function may be provided to perform the optimum correction before or after the beamforming process by adapting to the object. In these processes, although the degree of freedom is low, there are cases where analog processing is performed by analog devices or circuits instead of digital processing, emphasizing high-speed processing.

In the above processing, the superposition and spectral frequency division were linear processing, but nonlinear processing is performed after generation of waves by beamforming in the above methods (1) to (6). Generating a signal having a wave parameter is performed. In the beamforming process, when the received signal is an analog signal, it is based on analog signal processing using analog circuits (diodes, transistors, amplifiers, dedicated nonlinear circuits, etc.), and when it is a digital signal, it is digital using a digital signal processing unit. Based on the signal processing, the received signal may undergo exponentiation, multiplication, or other non-linear processing. Non-linear processing may be applied to the spectrum in the frequency domain.

In addition, as a modification of DAS, DAM (Delay and Multiplication) processing, which is the invention of the inventor of the present application, may be implemented in the frequency domain in the device of the present invention. Calculation of products such as exponentiation and multiplication in the spatio-temporal domain can be performed by convolution integral in the frequency domain. Signals can be frequency-enhanced, broadbanded, harmonically simulated, and steered waves can be omnidirectionally detected from at least one direction, e.g., the resulting wave images can be generated and it may be possible to measure displacement vectors using conventional unidirectional displacement measurement techniques.

It is also possible to generate imaging signals using virtual sources. Regarding the virtual source, there have been reports in the past that a virtual source is installed in front of the physical aperture and that a virtual source is installed at the transmission focal position. can be installed at any position, and the physical source and detector of wave motion can be installed at any position on an appropriate scatterer or diffraction grating (Patent Document 7, Non-Patent Document 8). The invention can also be implemented in those virtual sources and virtual detectors, as described above. It is possible to increase the spatial resolution and widen the field of view (FOV). In addition, when transmitting or receiving a received signal obtained by transmission using an aperture of one or more elements (with or without beamforming, transmission and reception for aperture plane synthesis), or performing beamforming for transmission and reception , a plurality of wave parameters, a plurality of beamforming parameters, and a plurality of transducer parameters (element shape, size, arrangement, number, effective aperture width, element material, etc.). By doing so, it is possible to generate multiple beams or waves with different characteristics (including obtaining multiple beams from the same received signal) and configure an over-determined system. By changing the position and distribution (shape, size, etc.) of , it is possible to construct an over-determined system as well. Again, the same received signal may generate multiple beams or waves with different characteristics. The high SN ratio and high resolution due to coherent superposition of these, which are the characteristics of overdetermined systems, and the reduction of speckle due to incoherent superposition through their detection may be effective in imaging. Also, the effect of improving the accuracy of various measurements such as displacement measurement and temperature measurement can be obtained. At least one of wave parameters, beamforming parameters, and transducer parameters may be different (e.g., electro-electronic or mechanical physically or software to change the steering angle, etc.).

An arbitrary wave source causes waves to be represented by rotating or spatially shifting the coordinate system determined by the receive aperture element array around an arbitrary position (e.g., the axis and lateral coordinates of the transmit aperture). are different from those determined by the reception aperture), the received signal may be subjected to coordinate correction before beam forming. For example, when it is desired to directly generate an imaging signal in a state in which the two-dimensional Cartesian coordinate system (x, y) is rotated by θ around the origin, the analytic signal obtained by the first Fourier transform in the time direction is can be calculated by multiplying the equation (29), rather than calculating the rotation of the wave vector (kx, √(k ² -kx ² )) and the coordinates (x, y) and the Jacobian, Image signals can be generated at high speed without losing the high speed achieved by the present invention.

However, in the case of the reflected wave, s=2, and in the case of the transmitted wave, s=1. Substantially, correction for transmission may be performed with s=1. Spatial shifting (translation) can also be performed in the frequency domain by multiplying by a complex exponential function. The above method of rotating the wavevector (kx, √(k ² -kx ² )) and coordinates (x, y) and calculating the Jacobian is similar to the transmit beamforming with transformation to the coordinate system determined by the receive aperture element array. (s=1), under which circumstances receive beamforming (s=1) is performed and is slow.

In the active type device according to this embodiment, the same applies to the passive type device according to the second embodiment described later, but as an analog device, it may be incorporated in a transducer or device main body. Lenses, reflectors (mirrors), scatterers, deflectors, polarizers, polarizers, absorbers (attenuators), multipliers, conjugators, phase delay devices, adders, differentiators, integrators, matchers, Special devices such as filters (space or time, frequency), diffraction gratings, spectroscopes, collimators, splitters, directional couplers, nonlinear media, wave amplifiers, etc. may be used together. In addition, when targeting light, polarizing filters, ND filters, shields, optical waveguides, optical fibers, optical Kerr effect devices, nonlinear optical fibers, optical mixing optical fibers, optical fibers for modulation, optical confinement devices, optical memory, dispersion shift Encoding by optical fiber, bandpass filter, time inverter, or optical mask, etc., and optical node technology, optical cross-connect (OXC), optical control (wavelength conversion, switching, routing) for them Optical add-drop multiplexers (OADMs), optical multiplexers/demultiplexers, or optical switch elements may be used, and the devices themselves may be optical transport networks or optical networks. This is not the case. In beamforming, they are to be controlled artificially, naturally, or equally optimally under the scheme as described above, together with the device. Non-linear processing may be applied to the spectrum in the frequency domain.

Under such combination, the device according to the invention can also be used in conventional devices using waves. Examples of medical equipment include ultrasonic diagnostic equipment (reflection/echo method, transmission type, etc.), X-ray CT (in some cases, a contrast agent that enhances the attenuation effect is used), X-ray X-ray, Angiography, Mammography, MRI (Magnetic Resonance Imaging, contrast agents may be used), OCT (Optical Coherent Tomography), PET (Positron Emission Tomography, corresponding to the second embodiment), SPECT (Single Photon Emission Computed) tomography), endoscopes (capsule types are also available), laparoscopes, catheters equipped with various sensing functions, terahertz imaging devices, various microscopes, various radiotherapy devices (chemotherapy may be used in combination to enhance therapeutic effects) ), a SQUID meter, an electroencephalograph, an electrocardiograph, and HIFU (High Intensity Focus Ultrasound). Among them, a nuclear magnetic resonance imaging apparatus is originally a digital apparatus, and has a very wide range of applications including its capabilities. For example, using the electromagnetic wave observation and inverse problem that the inventor of the present application is working on, it is possible to reconstruct current distribution and electrical properties (measurement), observe displacement and mechanical wave propagation, reconstruct mechanical properties (measurement), temperature It can also be applied to all of observation of distribution and heat waves, and reconstruction (measurement) of thermophysical properties (Patent Documents 8 to 10). Patent Documents 8 to 10 disclose a method that can reconstruct the distribution of physical properties related to one physical quantity (observed quantity) in a one-dimensional, two-dimensional, or three-dimensional region of interest. Observed physical quantities are surface waves (electromagnetic waves, elastic waves, thermal waves, etc.) and boundary physical quantities, and their physical properties can be similarly reconstructed based on the observations. When terahertz is used, it is possible to observe an electric field, and it is possible to observe current density, electric physical properties, and the like. Doppler observation using terahertz is also important. In addition, the distribution of physical properties may be evaluated directly from the waves themselves that are directly sensed by sensors. In regard to these, in addition to nuclear magnetic resonance imaging, efforts are being made to use ultrasonic waves, etc., and the same is true. In addition, as other efforts, for example, in OCT, it is possible to measure absorption spectra and image basal cell carcinoma of the skin, oxygen and glucose concentration in blood, etc. based on infrared spectroscopy. become. Although application to normal Near InfraRed (NIR) is also possible, it is possible to capture the distribution with higher spatial resolution than reconstruction based on so-called NIR. In these methods, an ultrasonic sensor device (a microscope is also possible) may be used in combination with an OCT or laser device to perform photoacoustic imaging, and applications are not limited thereto. In addition, lasers and OCT equipment may be used to capture and image tissue fluctuations in the absence of mechanical stimulation with high sensitivity (observation and application of surface waves such as Doppler observation). there is also). In addition, the response to any (mechanical) stimulus, including that caused by laser light, may be the object of imaging (including observation of dynamics generated using light, etc.). In other imaging, chemical sensors and the like may be used. Combinations of waves are not limited to these, and other sensors such as chemical sensors may be used in addition to physical sensors. Devices according to the present invention are also used in various radars, sonars, optical devices, and the like. Waves are not limited to pulse waves and burst waves, and continuous waves may also be used. In addition, digital processing with a high degree of freedom may be realized and used by a dedicated analog circuit with a fast operating time, and vice versa. Various devices exist in various fields such as resource exploration, non-destructive inspection, and communication, and the device according to the present invention is also used in these fields. The device of the present invention can be used in conventional devices as a device and in terms of modes of operation (e.g., imaging mode, Doppler mode, measurement mode, communication mode, etc.) It is not limited.

When arbitrary beams and waves such as multiple fixed focusing beams, multifocusing, and other plane waves are physically transmitted at the same time, high frame rates can be achieved if they can be transmitted over a wide area at once to the region of interest. realizable. In some cases, the signals are simultaneously transmitted in multiple directions from the same effective aperture, and in other cases, the signals are simultaneously transmitted in the same direction or different directions from different effective apertures. Such steering angles, focus positions, etc. may be the same or different in them, as well as ultrasonic frequency, bandwidth (beam direction, wave propagation direction, direction orthogonal to them), pulse shape, wave number, aperture shape and apodization. The same or different beam forming parameters such as beam shape, etc., and transducer parameters such as element shape, size, and arrangement may be transmitted at the same time.
In physical multiple transmission, when those parameters are different, the following can be considered as typical cases.
(A1) When applying the same soft steering to all.
(A2) When a plurality of different soft steerings are applied (for example, when the same steering is softly applied for each different transmission steering angle).
Also, when those parameters are the same, the following can be considered as representative cases.
(B1) When soft steering is applied.
(B2) When applying a plurality of soft steerings.
However, they may be implemented in combination. Obstacles in the propagation process and the effects of scattering and attenuation (including frequency-dependent cases) may lead to so-called adaptive beamforming. At that time, when the same combination of soft transmit and receive steering and apodization is performed, they are processed at once in a superimposed state. Also, if the combinations include different ones, each combination is processed once in a superimposed state for each identical combination to be performed, and superimposed before the final inverse Fourier transform.
In the cases of (A1) and (B1), one soft processing can be applied to the received echo signal received as a superposition of echo signals for each transmitted ultrasonic wave.
In the case of (A2), the echo signals are classified into echo signals to be subjected to the same processing in terms of software, each echo signal subjected to the same steering in terms of software is superimposed, each is subjected to the same software processing once, and finally. superimpose them before the inverse Fourier transform of . It should be noted that the signal separation may be performed using the various methods described above, and is not limited to them.
In case (B2), we apply different soft treatments to all superposed angular spectra and superimpose them before the final inverse Fourier transform.
In addition, when the beams and waves are not physically transmitted at the same time, the time phase of the target is the same, or under the assumption that multiple transmissions are performed, the above simultaneous It may be processed according to the transmission case. Also, when the same combination of soft transmit and receive steering and apodization is performed, they are superimposed and processed at once, and when different combinations are included, they are performed. Identical combinations are superimposed, each processed at once, and superimposed before the final inverse Fourier transform. Simultaneously transmitted and non-simultaneously transmitted may be treated similarly under the same conditions or assumptions above. These parameters in physical transmission may be known in advance, or beams and waves may be analyzed and used. This is also true in the case of the passive type, which will be described later.
By transmitting multiple beams and waves simultaneously or not, high frame rates and focusing on the same or multiple points can be realized. can be generated (for example, high resolution by widening the bandwidth), and also by frequency division processing of the spectrum, similarly, beams and waves with new parameters can be generated, and the beams and waves generated by them can be used as parameters may be used separately (for example, to measure the displacement in the direction of the generated beam or the propagation direction of the wave, or to measure the displacement vector). Band widening based on non-linear or linear processing, which will be described later, may be performed on the superimposed state, spectral frequency division, or separation. In addition, superposition, spectral frequency division, separation, non-linear or linear processing, etc. may be used for displacement measurement or the like. Each of them may be detected and imaged, or the detected ones may be superimposed and imaged (eg, to reduce speckle). Applications are not limited to these, and are diverse as described above, and are not limited to these.

<Simulation result>
In the following, the feasibility of the above beamforming methods (1) to (7) will be confirmed by simulation when the waves are mainly ultrasonic waves (plane wave transmission and monostatic aperture during steering). Image signal generation by surface synthesis, multi-static aperture surface synthesis, fixed focusing, image signal generation in the Cartesian coordinate system during transmission and reception in the polar coordinate system, and the result of the migration method).

FIG. 21 is a diagram showing a numerical phantom used in the simulation. Here, we dealt with a numerical phantom containing five point scatterers existing at 2.5 mm intervals at a depth of 30 mm in an anechoic and non-attenuating medium. Field II (see Non-Patent Document 21) was used to generate echo signals. Here, the depth direction is the z-axis, and the horizontal direction is the x-axis.

A one-dimensional linear array transducer (128 elements, element width of 0.1 mm, gap of 0.025 mm, thickness in the slice direction of 5 mm) was used in plane wave transmission, migration method, and monostatic aperture synthesis. A one-dimensional linear array type transducer (256 elements, element width of 0.1 mm, gap of 0.025 mm, thickness in the slice direction of 5 mm, effective aperture width of 33 to 129 elements) was used for fixed focusing and multi-static aperture plane synthesis. A convex transducer (128 elements, element width of 0.1 mm, gap of 0.025 mm, thickness in the slice direction of 5 mm, radius of curvature of 30 mm) was used for transmission and reception in the polar coordinate system. The center frequency of the emitted ultrasonic pulse is set to 3 MHz, and its sound pressure waveform is shown in FIG. The deflection angle is defined with respect to the depth direction of the front surface, and is hereinafter represented as "θ".

(1) Transmission of polarized plane waves Simulation results when transmitting polarized plane waves with deflection angles θ = 0°, 5°, 10°, and 15° are shown in Fig. 23A ((a) to (d), respectively). . FIG. 23B shows a simulation result when interpolation approximation is performed in wave number matching when the deflection angle is θ=0°. These are the results of receiving at the same angle. In FIGS. 23A and 23B, the horizontal axis represents the position [mm] in the horizontal direction (Lateral x), and the vertical axis represents the depth (Depth) z [mm]. As shown in FIG. 23A, a focused echo image is obtained and can be confirmed to be deflected. Both are the results of down-sampling from 100 MHz to 10 MHz in the frequency domain (paragraphs 0208 and 0209) and generating image signals at intervals corresponding to the sampling frequency of 25 MHz (same for others).
On the other hand, when the deflection angle is θ = 0°, interpolation approximation is performed in wave number matching, and Fig. 23B (e) shows sampling frequencies of 100 MHz (left) and 25 MHz (right) when the spectrum is replaced with the neighboring spectrum. FIG. 23B(f) shows the result when the sampling frequency is 100 MHz (left) and 25 MHz (right) when linear interpolation approximation is applied to the spectrum. The image was more stable when the sampling frequency was higher and linear interpolation approximation was performed, but it was not as good as FIG. 23A(a) when interpolation approximation was not performed. A non-zero deflection angle gave even more unstable results.

24 and 25 show the results of calculating the obtained deflection angle from the spectrum of the generated image signal. For this evaluation, the number of scatterers in the numerical phantom was increased, and 300 scatterers were randomly placed at depths from 0 to 40 mm. The reflection coefficient of each scatterer was a random value between -1 and 1. An error of 0.5 to 0.8 degrees was confirmed regardless of the set deflection angle. The error is dependent on the scatterer's position relative to the generated wave. Accuracy improves with more scatterers (omitted).

FIG. 26 shows the result of superimposing a plurality of images obtained at different deflection angles. Deflection angles are set at intervals of 1°, 1 wave (0°), 11 waves (-5° to 5°), 21 waves (-10° to 10°), 41 waves (-20° to 20°). superimposed on each other. FIG. 27 plots the lateral distribution of the point spread function estimated from the generated image signal. In FIG. 27, the horizontal axis represents the position [mm] in the horizontal direction (Lateral x), and the vertical axis represents the relative value of brightness. As shown in FIG. 27, it can be confirmed that the resolution improves as the number of superpositions increases.

Migration method (applying method (6) to beamforming when transmitting a plane wave with the same polarization)
FIG. 28 shows the result of generating an image signal by the migration method according to the present invention. The deflection angles were set to θ=0°, 5°, 10°, and 15° as in FIG. 23A. In the wavenumber matching, the image was unstable when interpolation approximation was applied.

(2) Deflection Monostatic Synthesis of Aperture Planes FIG. 29 shows simulation results of deflection by monostatic aperture plane synthesis of method (2). Similar to FIG. 23A, the deflection angles were set to θ=0°, 5°, 10°, and 15°. As shown in FIG. 29, it can be confirmed that an image is formed and deflected.

(3) Multistatic Aperture Synthesis FIG. 30 shows the results of simulation by multistatic aperture synthesis of method (3). FIG. 30(a) shows a low resolution image generated using only the received signals (ie one set) with the receiving elements equal to the transmitting elements (ie same as monostatic). FIG. 30(b) shows the result of superimposing the results of a total of 33 sets using the left and right 16 elements for reception in addition to the element at the transmission position. FIGS. 30(c) and (d) respectively show the result of superimposing 65 sets (32 left and right transmitting elements) and the result of superimposing 129 sets (64 left and right transmitting elements). As shown in FIG. 30, it can be confirmed that an image has been formed. FIG. 31 shows the results of plotting the respective point spread functions. From FIGS. 30 and 31, it can be confirmed that the side lobe is suppressed as the number of overlaps increases, and the resolution is high.

(4) Fixed Focusing FIG. 32 shows the result of fixed focusing transmission. Here, method (1) was used. FIG. 32(a) shows the result of superimposing the echo signals received at each transmission effective aperture and performing one echo signal generation process. FIG. 32(b) shows the result of generating low-resolution echo signals for each effective aperture width and superimposing them. FIG. 32(c) shows the result of generating and superimposing echo signals by using a set of signals having the same positional relationship for transmission and reception as in the case of multi-static aperture plane synthesis. As shown in FIG. 32, images are formed in either case, and there is no particular difference. Method (1) is computationally faster and more efficient than the other two methods. In addition, the results of method (1) show that reception beamforming is possible even when ideally received signals are used when multiple or all beams are simultaneously transmitted (for transmission beams with interference It can also process received signals and achieve high-speed frame rates). This processing is not limited to fixed focusing transmission, and can be performed during multiple transmissions of all kinds of waves (including combinations of different waves). In other words, there are cases where multiple waves include those with different transmission beamforming, cases with and without beamforming, and cases with different types of waves (electromagnetic waves, mechanical waves, heat waves, etc.). However, in some cases, processed data such as non-linear processing, detection processing, super-resolution processing, adaptive beam forming, minimum variance processing, signal separation, and the like are included. These processes may be performed during beamforming. Of course, the received signals for each transmission may also be superimposed. Interpolation approximation processing may also be performed in wave number matching in these cases.

(5) Generation of image signals in the Cartesian coordinate system of transmission and reception in the polar coordinate system (5-1) Transmission of cylindrical waves Simultaneously radiating ultrasonic waves from all elements of the convex array and transmitting the cylindrical waves FIG. 33(a) shows the result of processing by region. Actually, as described in (5-1′), plane waves or virtual linear arrays are generated at a depth of 30 mm for transmission and reception. As shown in FIG. 33(a), it can be confirmed that the scatterer is imaged.
(5-1') Cylindrical wave transmission using a linear array Next, an echo signal when a cylindrical wave is transmitted using a linear array and a virtual sound source (Fig. 8A (a)) set at a position behind it shows the result of generating FIG. 33(b) shows the result of using the method (1) described in the method (5-1′) with the virtual sound source positioned 30 mm behind, and FIG. It shows the result of using the method (2) described in the method (5-1′) at a position 60 mm behind. It can be confirmed that the scatterer is imaged.
In addition, in the case of using a linear array type transducer, the method (1) described in the method (5-1′) is used to generate a cylindrical wave using a virtual source located 30 mm behind the physical aperture. 33(d) shows the result of generating a laterally spread plane wave or a virtual linear array type transducer (FIG. 8B(g)) at a distance of 30 mm.

(5-2) Fixed Focusing Using a convex array, fixed focusing is performed at a distance of 30 mm from each element (FIG. 14(a)). The results of processing received signals are shown in FIGS. 34(a) and 29(b). . FIG. 34(a) shows the result of superimposing received signals obtained at each effective transmission aperture and performing one echo signal generation process, and FIG. The results of generating and superimposing images are shown. Similar to the multi-static aperture plane synthesis, echo signals are generated with the same positional relationship for transmission and reception as a set, and the result of superimposition is omitted, but these three calculations are almost the same as in (4). brought about the result of Also, FIG. 34(c) shows the result of performing one echo signal generation process on the received signal when fixed focusing is performed at a depth of 30 mm (FIG. 14(b)). The scatterers were well imaged.
These results were obtained in the same manner as in (4), and indicate that reception beamforming is possible even when ideally received signals are used when multiple or all beams are simultaneously transmitted. (It can also process received signals for transmit beams with interference and achieve high frame rates). This processing is not limited to these fixed focusing transmissions, and can be performed during multiple transmissions of all kinds of waves (including combinations of different waves). In other words, there are cases where multiple waves include those with different transmission beamforming, cases with and without beamforming, and cases with different types of waves (electromagnetic waves, mechanical waves, heat waves, etc.). , non-linear processing, detection processing, super-resolution, adaptive beam forming, minimum variance processing, signal separation, and other processing may be included. These processes may be performed during beamforming. Of course, the received signals for each transmission may also be superimposed. Interpolation approximation processing may also be performed in wave number matching in these cases.

The beamforming using the digital Fourier transform according to the present invention performed in the above simulation is based on the appropriate use of multiplication of complex exponential functions and Jacobi arithmetic, and arbitrary beamforming processing in an arbitrary orthogonal coordinate system is performed. We have demonstrated that it can be realized with high accuracy without interpolation approximation. Both beamforming can be realized using the DAS (Delay and Summation) method. If a PC is used, the calculation time is significantly faster by a factor of 100 or more. If the aperture elements form a two-dimensional or three-dimensional distribution, two-dimensional or three-dimensional array, then the above method can be made multi-dimensional, requiring even more processing than the one-dimensional case. It solves the problem of requiring more, and its high speed becomes even more effective. Also, an embodiment has been described in which superimposition and the like are effective when plane waves are transmitted with different deflection angles. It is possible to obtain effects such as high resolution and suppression of side lobes and high contrast at high speed.

In the above example, it is confirmed that arbitrary focus (including no focus) and steering can be performed in arbitrary array-type aperture shapes, and arbitrary beamforming processing in arbitrary orthogonal coordinate systems can be performed at high speed and without interpolation approximation. It was confirmed that this can be achieved with accuracy. The time required to obtain high-order measurement results such as displacement measurement based on the generated image signal is also shortened, and the effect of improving the measurement accuracy is also obtained. In the present invention, as described in methods (1) to (7), interpolation approximation processing may be performed in arbitrary beamforming wave number matching, and beamforming may be performed at a higher speed. In order to perform approximate wavenumber matching with high accuracy, it is necessary to appropriately oversample the received signal at the expense of an increase in the amount of calculation. In that case, it should be noted that the number of data for Fourier transform increases unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed.

An example using the representative transducer, receiving sensor, transmitting unit and receiving unit, control unit, output device, external storage device, etc. in the first embodiment has been described above. The fact that the beamforming of methods (1) to (7) was possible means that it is possible to perform arbitrary beamforming based on focusing and steering in an arbitrary orthogonal coordinate system. Beamforming and applications that can be realized using , including but not limited to those otherwise described.

<<Second Embodiment>>
Next, the configuration of the measurement imaging apparatus or communication apparatus according to the second embodiment of the present invention will be described. FIG. 1 is a representative block diagram showing a configuration example when the device according to the first embodiment is of the active type, and FIG. 2 is a representative block diagram showing in detail a configuration example of the device main body. However, in the second embodiment, a passive device is used, so the device according to the second embodiment does not comprise at least a transmitting transducer in FIG. It does not provide a wired or wireless path for sending drive signals to the transducer.

In the case of the active type device according to the first embodiment, detailed configuration examples of the device and units have been described with reference to FIGS. 1 and 2 as representative configurations. Passive devices use arbitrary aperture shape transmitting transducer array devices in them, whereas passive devices necessarily use transmitting and receiving transducer array devices (a transducer may be used for both transmitting and receiving). Do not use

That is, the basic configuration of the device according to the second embodiment includes a receiving transducer (or receiving sensor) 20 as a receiving means, a device body 30, an input device 40, an output device (or display device) 50, and an external and a storage device 60 . The device body 30 mainly includes a receiving unit 32, a digital signal processing unit 33, a control unit 34, and a storage unit (memory, storage device, or storage medium) not shown. Furthermore, the device body 30 may include a transmission unit 31 . The description of these components in the first embodiment also applies to the second embodiment.

As in the first embodiment, each of these devices and each unit in the device main body 30 can be installed at a remote location. The device main body 30 is composed of a plurality of such units, and is so called for convenience. Further, as in the first embodiment, reception may be performed by mechanically scanning the receiving transducer 20 . Even if it is not generally called an array type, it may be processed in the same way.

However, unlike the device of the first embodiment, the device of the second embodiment is designed to sense when the waves are generated, as will be explained in detail below, from arbitrary wave sources. A timing signal is generated by receiving the wave itself of the object to be observed, or a timing signal generated through another process is sensed by the control unit through wired or wireless communication, and the data of the receiving unit is captured (each receiving channel AD conversion and writing to memory).

The receiving transducer (or receiving sensor) of the device according to the present invention, when the wave itself coming from the wave source is used as the timing signal as a way for the control unit to sense the timing signal that tells when the wave is generated. Either the received signal itself received by the receive aperture element 20a of 20 is used, or a timing signal received by a dedicated receiver that may be provided in the device body 30 is used.

In this case, the reception aperture element 20a (all elements may be used, but in the physical aperture, elements at the end or central position may be sparsely used) or a dedicated reception device (at least a plurality of reception channels may be used). ) is continuously detected in time, and information about the signal strength, frequency, band, or code of the received signal is obtained, for example, through various input means as described above. is set in the control unit 34 itself (built-in memory) or analog judgment circuit (in this case, it is not variable in terms of software or hardware, and may be fixed in terms of hardware). Alternatively, the sensing of the timing at which waves are generated can be obtained by translating the signal received by the receive aperture element 20a or a dedicated receiver into a threshold or value recorded in a memory or storage device (storage medium), a characteristic of the wave to be observed. Based on matching with discrimination data such as a database related to.

When the received signal is determined in an analog manner, a dedicated analog circuit is provided for determination, and only when the signal is determined to be the object of observation, a trigger signal for data acquisition is generated, It may be converted and stored in a memory or storage device (storage medium), and beam forming processing may be performed.

When the received signal is digitally discriminated, the received signal is AD-converted continuously over time and stored in a memory or a storage device (storage medium), and is always or at any time (when a command is given through an input device, etc.). ), at a predetermined time interval (set through an input device, etc.), the digital signal processing unit 33 reads out the stored signal, compares it with the same discrimination data, discriminates it, and determines it as the signal to be observed. Beamforming processing may be performed only when

Since the storage capacity of memory and storage devices (storage media) is finite, in the case of digital discrimination, the wave signal of the observation target was not observed within a predetermined time (set through an input device, etc.) In this case, the memory address is initialized. In addition, although it is not efficient in terms of energy saving, beam forming may be performed as needed, and wave signals may be discriminated using the same discriminant data based on image signals whose accuracy has been improved by beam forming. The process is similar when waves for normal communication purposes are observed.

In addition, the dedicated receiver may be installed in another location away from the units of other devices, such as a location near the wave source to be measured or a location with a good reception environment for the timing signal. In some cases, a wave that propagates faster than the wave received by the reception aperture (a wave that becomes a timing signal) is used, and the timing signal is transmitted to the control unit in the main body of the device via the dedicated receiver. . Dedicated lines (wired or wireless) that may use relay stations may also be used. In this case, the timing signal is used as a trigger signal to capture the received signal (AD conversion and storage in a memory, storage device, or storage medium) and beam forming.

Also, the timing signal from which the wave was generated may arrive after the device of the present invention receives the wave. In other words, the propagation speed is slow, or such a mechanism may be used, and it may be so as a result. In order to cope with such a case, the reception signal is continuously captured at all times, and the reception signal stored in the memory or storage device (storage medium) is read back in time and beamforming is performed. . In this case, information about waves obtained by other observers or observation devices is added to the timing signal as additional information at a relay station or the like, and the timing signal with the additional information is transmitted and dedicated reception is performed. Information including additional information may be read by the device and used in other devices as well as the device of the present invention. The line used is not limited to a dedicated line, and a normal network may be used. Similar timing signals may also be used when waves for normal communication purposes are to be observed. Additional information may be conveyed in a wave or signal separate from the timing signal.

Also, along with the generation of the wave to be observed, the wave that propagates at a higher speed or slower than the wave received by the receiving aperture element before, during, or after the generation (the wave that becomes the timing signal) is the wave source. , and their dedicated receivers and dedicated lines may be installed and used as well. In that case, information about the wave to be observed may be added to the wave that becomes the timing signal, and information about the wave obtained by another observer or observation device may be added by a relay station and transmitted. and the information, including additional information, may be read by dedicated receivers and used by the apparatus of the present invention and other devices. The line used is not limited to a dedicated line, and a normal network may be used. Similar timing signals may also be used when waves for normal communication purposes are to be observed. Additional information may be conveyed in a wave or signal separate from the timing signal.

As these dedicated receiving devices, dedicated sensing devices capable of sensing timing signals or reading additional information are used. Any active or passive observer or similar observer of waves, or any other device that is a precursor to the wave's production, or is produced simultaneously with the wave's production, or separates after the wave's production. Any active or passive observer or similar observer of the phenomenon or wave of Alternately, the dedicated receiving device may only receive the timing signal, and the reading of the additional information itself may be performed by a digital signal processing unit via a dedicated device or a control unit within the device itself.

Additional information may also be read by the digital signal processing unit 33 when the active or passive device of the present invention itself is used as a sensing device. Timing signals may also be generated by provided sensing devices. Efficiency of data acquisition operation and beamforming process, power saving, memory and storage devices when it is not known when, where, or when and where the wave will be generated It is important for saving (storage media). Data acquisition and beamforming are performed based on the clock signal of the control unit 34 . If the wave source is digital, it is better to be synchronous, and based on the digital reception of the wave to be observed, the device may run at high clock and sampling frequencies. The same is true when the timing signal is an analog signal, but when the timing signal is a digital signal, the apparatus itself may synchronize.

The object of observation is the wave itself generated by a self-emanating wave source, and the characteristics of the wave source (strength, type of source, etc.), position, time of activity as a wave source, etc. is sometimes observed. In addition, in the same way as when the device is of the active type, the distribution of temperature (distribution), displacement, velocity, acceleration, strain, strain rate, etc. of the object may be obtained from the wave spectrum. In addition, the characteristics of the medium in the propagation process (propagation velocity, physical properties related to waves, attenuation, scattering, transmission, reflection, refraction, etc., or their frequency dispersion, etc.) can be observed, and the structure and composition of the observation target and medium is sometimes revealed. For example, radioactive materials (such as isotopes in PET), materials with non-zero thermodynamic temperatures, seismic sources, neural activity, astronomical observations, weather, incoming objects, moving objects, communication equipment including mobile communication equipment, physical Alternatively, objects that react to chemical stimuli, electric sources, magnetic sources, radiation sources, various energy sources, etc. are observed, and observation targets are not limited to these.

Fusion of measurements and data mining may also be done through multiphysics or multichemomics using multiple different types of wave receiving transducers or receiving sensors. Of course, they can also be observed by a single transducer or sensor (e.g., in medical ultrasound imaging, strain and blood flow, which represent tissue deformation, and their related tissue properties can be selectively At the same time, different colors are superimposed on the echo image, and different markers in a wide band are imaged with different colors at the same time. display with different colors, or image with different color densities, etc.). For those that function based on multiple functions or many physical properties, or those that affect the surroundings in different ways, multifaceted behavior of the whole (whole body in the case of humans) and local behavior It is also done to understand the whole (whole body in the case of humans) and local behavior of the object newly or in detail. For example, various neural controls (body temperature, blood flow, metabolism, etc.) that are performed in a short time or over a long time in living things, short-term or long-term effects on living things (radiation It can be used for the development of artificial organs, cultured tissues, their hybrids, medicines, supplements, etc. that contribute to longevity and life extension, and for monitoring their actions.
Japan and other developed countries are facing an aging society, so it is important to improve QOL and reduce medical expenses. Cancer lesions such as human liver cancer, pancreatic cancer, kidney cancer, thyroid cancer, prostate cancer, breast cancer, etc. are increasing (more than 1.5 million people in Japan, 1 out of 2 people are affected, and 1 out of 3 people are the cause of death) and uterine fibroids, brain diseases (brain tumor, cerebral hemorrhage, cerebral infarction, more than 1.3 million people), ischemic heart disease (myocardial infarction, angina pectoris, more than 1.7 million people), arteriosclerosis/thrombosis (1 in 4 people) Cause of death), dyslipidemia (over 2.06 million people), diabetes (more than 10 million people), and other adult-onset diseases. However, the present invention opens up innovative integrated diagnostic imaging systems and techniques and clinical styles (including screening) that enable early non-invasive differential diagnosis to be performed with high accuracy, convenience, and low cost.
Most of cancer metastases are metastasis to lymph nodes where lymph flow gathers (lymphatic metastasis), and metastasis to places with abundant blood flow such as lung, liver, brain, and bones (hematogenous metastasis). . Currently, regarding lymphoma, when the swelling or lump is first diagnosed, the spread (disease stage) and general condition are diagnosed. In addition, through the bloodstream, for example, pancreatic cancer to the liver, peritoneum, etc., breast cancer to the liver, lung, brain, etc., gastric cancer to the lung, liver, kidney, pancreas, etc., lung cancer to the liver, kidney, brain, etc. It is known that colorectal cancer metastasizes between organs such as the liver, lungs, and brain. Because metastatic cancer has the characteristics of the primary cancer, it is imperative to identify the primary cancer for appropriate treatment. Of course, it is also important to understand the hepatitis, chronic viral hepatitis, liver cirrhosis, etc. (the mechanism of development itself) that cause it, for example, as in the case of primary hepatocellular carcinoma. In addition, there is a high possibility of prostate disease and uterine fibroids if there is a large amount of blood flow in the feeding arteries around early cancer tumors and advanced cancer tumors, and blood flow is poor. and hyperlipidemia, blood sugar viscosity is high, the risk of developing cancer is 1.2 to 1.3 times higher with or without a history of diabetes, and the risk of developing ischemic heart disease is three times higher in the diabetic group. Dyslipidemia (20-50% of diabetes mellitus) and their promotion of arteriosclerosis, increase in blood flow due to increased temperature due to heating (perfusion), and kidney regulation of red blood cell count (oxygen content) and blood pressure. Therefore, the present invention is important for in situ observation of pathological conditions, neural control, and hemodynamics of these various tissues simultaneously in real time and with high accuracy. As for thrombosis, it is also desirable to improve the accuracy of observation of hemodynamics after surgery and in the daily life of cardiac pacemaker users (welfare).
For example, under the above approach that focuses on the network of nerves, blood flow, and lymph, specifically MRI (magnetic resonance imaging), SQUID (superconducting quantum interference device), (optical) ultrasound device, OCT (optical coherence tomography) A new high-precision and simple real-time three-dimensional in vivo system based on the mathematical inverse problem of three basic physics, electromagnetics, mechanics, and thermodynamics (described in paragraph 0377, etc.) Using diagnostic imaging techniques, it is possible to implement early integrated imaging techniques for the properties of those related organs and tissues (related simultaneous observation and fusion imaging by simultaneous multi-observation of the same or multiple organs). As a minimally invasive treatment for heart disease, it can be combined with HIFU (High Intensity Focus Ultrasound), electromagnetic heat coagulation treatment, chemotherapy, pharmacological therapy, and various radiotherapy. Accurate and inexpensive early differential diagnosis and minimally invasive early treatment can be performed. It opens up an innovative clinical style that allows rapid early diagnosis followed by heat treatment in the shortest possible time.
(1) <Electromagnetism-based> Imaging of current density vectors and electrical properties using MRI and SQUID: Based on magnetic field measurement, imaging of current density vectors and three-dimensional distribution of electrical properties (conductivity and permittivity) in cranial nerves. Visualize the network (superimposed on MRI diffusion imaging). The inverse problem of the Biot-Savart law and the differential inverse problem using the observed current density vector, respectively. For example, perfusion (blood flow) control can be monitored in fusion with HIFU heating treatment in (6) below.
(2) <Mechanics-based> Imaging of blood flow vector, blood pressure, and shear viscosity using MRI, (light) ultrasound, and OCT: In MRI, based on distribution imaging of hydrogen, carbon, phosphorus, etc., In the case of sonic echo method and photoacoustic ultrasound (under the distinction between arteries and veins, when using D-glucose or L-glucose, which is specifically taken up by cancer as reported recently, as a contrast agent, sugar due to diabetes, When LDL, HDL cholesterol and neutral fat due to dyslipidemia are used as markers), the three-dimensional distribution of the three-dimensional flow velocity vector and strain rate tensor of the blood flow is imaged using the multidimensional Doppler method (normal Doppler and Unlike , it is possible to observe vectors in any direction just by pointing the sensor at the target), and based on the differential inverse problem from the observed flow velocity data, the three-dimensional distribution of blood pressure, intracardiac pressure, shear viscosity, and density is imaged. MRI is intracranial cerebral blood flow, ultrasound echo method is intracardiac, abdominal organs (liver, pancreas, kidney, prostate, uterus), body surface tissues (breast, thyroid, eye, skin) Blood flow, optical ultrasound, and OCT are used to monitor blood flow in open body organs, eyes, and skin (eye monitoring in connection with diabetes). Finally observed the blood flow network. Diagnosis of ischemic heart disease (myocardial infarction, angina pectoris), brain disease (brain tumor, cerebral hemorrhage, cerebral infarction), arteriosclerosis, and thrombosis, focusing on dyslipidemia and blood sugar, by quantifying blood pressure, intracardiac pressure, and viscosity. Simultaneous observation of (3) soft tissue is performed in (5) integrated imaging. Or monitoring perfusion (blood flow) in HIFU heat treatment of (6).
(3) <Mechanics-based> Imaging of soft tissue dynamics and shear (visco)elasticity using MRI, (optical) ultrasound, and OCT: Ultrasound echo method and optical ultrasound (focusing on soft tissue components, innovative In the case of broadband observation of irradiated light and photoacoustic waves without contrast agents or markers), using the same multidimensional Doppler method as in (2), the three-dimensional distribution of the three-dimensional displacement vector and strain tensor of the soft tissue Furthermore, based on the differential inverse problem from the observed displacement data, the three factors of body pressure (tissue pressure and intraocular pressure), shear (visco)elasticity (hardness), density, and force source (radiation pressure and HIFU) In situ imaging of dimensional distribution. MRI is intracranial brain tissue and lymphatic network, ultrasound echo method is heart tissue (myocardium and each valve), abdominal organs (liver, pancreas, kidney, prostate, uterus), body surface tissue (breast, thyroid, eye, skin) , Lymphatic network, photoacoustic ultrasound, and OCT diagnose pathological conditions (tissue properties) of organs, eyes, and skin in an open body. Simultaneous observation of blood flow of (2) is performed in integrated imaging of (5). Alternatively, monitoring the effect of (6) HIFU heat treatment (denaturation such as tissue coagulation).
(4) <Thermology-based> Imaging of soft tissue temperature and thermophysical properties using MRI, (light) ultrasound, and OCT: In order to monitor metabolism and (6) HIFU heat treatment, so-called MRI Larmor frequency chemical In addition to observations using temperature dependence of sound velocity and volume of (optical) ultrasonic waves, in situ observation of three-dimensional temperature distribution using temperature dependence of shear (visco)elastic modulus using (3). We conducted robust and practical observations on tissue displacement. Furthermore, based on the differential inverse problem, thermophysical properties (thermal conductivity, heat capacity, thermal diffusivity), perfusion (blood flow observation in (2) can be applied), and 3D distribution of heat source (HIFU) are imaged. , (6) for heat treatment planning in HIFU heat treatment.
(5) <Integrated Imaging> Simultaneous multi-observation of multiple organs by (1)-(4): Based on three-dimensional imaging of tissue physical properties of electromagnetism, mechanics and thermology, focusing on the network of nerves and blood flow/perfusion Integrated imaging diagnosis that simultaneously observes multiple organs from multiple perspectives. It is easy to use and performs diagnosis in a short time.
(6) <Fusion of integrated imaging and treatment> HIFU heating treatment control (automatic treatment) using integrated imaging in (5): In order to complete treatment in a short time in a minimally invasive manner, blood flow data in (2) and ( 4) Observation data of temperature and thermophysical properties and HIFU heat source database are applied to a temperature distribution calculation simulator to predict temperature distribution, and monitoring data of treatment effect (denaturation) by shear (visco)elastic imaging of (3) is also utilized. Using an updated automated heating treatment plan, controlling the focal (heating) position of the HIFU beam, beam shape, beam intensity, exposure time, and exposure interval under phase aberration correction. The accuracy, reliability, and safety of treatment are improved, and the HIFU applicator can be installed in the vicinity without obstacles such as bones, making it possible to treat widely spread organ cancers other than prostate cancer and uterine fibroids.
It is desirable that the treatment means be minimally invasive, and as described above, the treatment is not limited to these. The present invention is a technique that is simple, inexpensive, and brings about a high quality of life in an aging society. (Theranosis). The above imaging and treatment can be performed not only from the body surface, but also during surgery (directly to the organ or through the organ while the body is open), during laparoscopic surgery, orally, through the nostrils, through the portal, and through the vagina. can also be implemented.
It also includes those that substitute for various sensors that living things have, those that supplement them, or those that have new sensors. In the case of living organisms, it may be necessary to make them compact and wearable, or to use shapes and materials that are familiar to living organisms. There are various processing contents, for example, when multiple compression waves and shear waves arrive at the same time as dynamic waves, analog dedicated devices are used using modes, frequencies, bands, signs, propagation directions, etc. Alternatively, a digital signal processing unit may be used to separate the waves as in the first embodiment before beamforming. When there are multiple wave sources of electromagnetic waves, multiple electromagnetic waves with different characteristics may be superimposed and separated. Alternatively, due to the phasing and addition effect of beamforming, even when a plurality of waves arrive, a highly accurate image signal may be generated (for example, when the medium is a scattering medium).

Of course, signals may be separated based on the same processing after beamforming is performed. In order to obtain the effect of phasing and addition, it is necessary to obtain the wave arrival direction and the position of the wave source. In reception, besides dynamic focusing, fixed focusing is also useful. In order to obtain them, the centroid (center) frequency and instantaneous frequency of the multidimensional spectrum of the wave received by the reception aperture element array, the band, so-called MIMO, SIMO, MUSIC, independent component analysis, code, or various parametric methods, etc. is sometimes used. In some cases, the same processing is performed after beamforming, but in other cases, waves are observed using geometric information, especially after beamforming at multiple positions. . The processing method is not limited to these, and may be implemented under an inverse problem approach, for example.

For example, the direction of propagation of incoming waves can be obtained based on multidimensional spectrum analysis of received signals (past achievements of the inventor of the present application). Alternatively, even if no information about the propagation time is obtained using the effective reception aperture (usually, the position and distance of the wave source are determined from the time when the wave was observed at multiple positions), the wave source can be geometrically determined. It is possible to calculate the position and distance. Waves can be observed not only as pulse waves and burst waves but also as continuous waves. If the arrival direction of the wave is known through any processing, it is also possible to steer and focus the receive beam in that direction (monostatic or multistatic aperture synthesis) and observe it in detail. If necessary, the active type of the first embodiment may be used to perform transmission beam forming. In those processes, receive beamforming is performed while changing the steering angle, always emphasizing the most likely direction, and the obtained image or imaging, spatial resolution, contrast, signal strength, etc. are observed, or Through multidimensional spectral analysis, the direction of the wave source can also be determined. It may be automatically controlled.

Super-resolution may increase the resolution of an image signal. There is also a description in paragraphs 0009 and 0425. It may be easier to measure the size, intensity, position, etc. of the wave source, the object to be measured, the scattering in the medium, and the reflector. Although the band is necessarily limited by the physically generated wave field, typical super-resolution is to widen the band by inverse filtering and restore them as the original signal source or signal. be. Also, waves are usually subject to frequency-dependent attenuation, can be unfocused, and the source of the wave can be a moving object, or it can create disturbances in the intervening medium. . Super-resolution is sometimes implemented to correct these. As described in paragraph 0383, it may be useful to apply more filtering to have the desired point spread function, rather than just broadening the bandwidth, and in various super-resolution, this filtering can be combined with inverse filtering to frequency Implementation in the domain or spatio-temporal domain is a feature of the invention.

In addition, the object to be measured may move during the transmission and/or reception required to generate one image signal, and motion compensation may be required. In many cases, the point spread function is unknown, and in such cases, blind deconvolution may be performed, including the case where the above signal separation processing (particularly, blind separation) is also used. The method described in paragraph 0425 and the like are known. In addition, there are various other methods such as the maximum likelihood method (for example, Non-Patent Documents 39 to 41, etc.). It is desirable to evaluate the point spread function by some method such as obtaining an autocorrelation function, and ideally obtain a coherent point spread function. is obtained, inverse filtering is possible. Even in such cases, it is useful to perform the filtering together to have the desired point spread function.

If the point spread function cannot be obtained when observation is desired, for example, the point spread function should be evaluated and stored as a database when observation is possible. One effective method of inverse filtering is to weight the observed spectrum so that it is identical to the desired point spread function, echo distribution, or other spectral distribution of the signal distribution (of course, the distribution of spectral intensity). is possible. Spectral distribution of signal distribution such as desired point spread function and echo distribution may be set analytically or through simulation, optimization, etc., and beamforming is performed under ideal parameters for the measurement target. , it may be done with a single measurement, ensemble averaging may be performed under multiple measurements, or addition averaging may be performed under the assumption of a local stationary process ( It is also found in a similar manner using a phantom (calibration phantom) of the object to be measured (for example, Non-Patent Documents 35 and 36), which has been performed for a long time. A one-dimensional point spread function in the wave propagation direction or in a direction orthogonal to it may be estimated by obtaining a one-dimensional autocorrelation function, but a multidimensional point spread function may be estimated by obtaining a multidimensional autocorrelation function. There is also (Non-Patent Documents 8 and 14). They are equivalent to one-dimensional and multi-dimensional spectra in the propagation direction and orthogonal direction, respectively (ie, self-spectrum). These are used when evaluating the wavelength of wave motion, the shape of the force source, and the spatial resolution of wave motion (Patent Document 11, etc.), and are used for super-resolution. For example, fixed focusing or aperture synthesis (Non-Patent Document 15) is performed using a power function type apodization to obtain a desired high-resolution point spread function or signal distribution such as an echo distribution, and a plane wave capable of high-speed transmission and reception. A low-resolution signal (Non-Patent Document 15) obtained by transmission using Gaussian apodization may be increased in resolution. The latter is suitable for high-precision measurement of high-speed motion and shear wave propagation, and can realize both measurement and high-resolution ultrasound imaging at the same time. Alternatively, inversion can be performed using the power spectrum of the signal itself. The angular spectrum before wavenumber matching or the spectrum after wavenumber matching can be subjected to these processes. That is, super-resolution can be performed using the angular spectrum or spectrum of a signal distribution such as a desired point spread function or echo distribution, or the angular spectrum or spectrum of a received signal. In weighting, when dividing by a zero value or a small spectrum, it is necessary to be careful. In particular, it is not effective to amplify various noises contained in the received signal. ), Wiener filter (suppresses amplification of frequency components with low signal-to-noise ratio), singular value decomposition (throw away small singular values and spectra and do not use signal components with corresponding frequencies), Likelihood estimation (with or without MAP) is effective.

Inverse filtering can also be performed in the processes of methods (1) to (7) in the above digital wave signal processing. The resolution of the image signal is increased, and the quantitative (numerical value) effect may also be obtained, and the same effect may be obtained even when displayed as an image. You can get the effect of restoring a blurred image or bringing it into focus. Inverse filtering may be performed on an incoherent signal, and when performed on a coherent signal, it is effective. In particular, spatial changes in physical property distribution may be understood. Super-resolution may also be applied to superimposed or frequency-divided spectra. Applications of super-resolution are not limited to these.

New super-resolution can also be implemented. One is based on the non-linear processing described below and the other is based on instantaneous phase imaging as described here.

であり、ｔ＝０は、t軸方向の基準位置、即ち、波動源の位置（t＝０）を表し、δθ(t)は、位置座標tにおいて反射や散乱により生じる位相変化を表す。
とすると、これより、伝搬方向tの瞬時角周波数ω(t)と、瞬時位相θ(t)等を求めてイメージングする。伝搬方向tは、ステアリングせずに正面方向を向いている場合もあるし、ステアリングして偏向角度を持つ場合もあり、関心領域が３次元であるときも、又、２次元、１次元であるときもある。非特許文献１９にある通り、波動又はビームの伝搬方向を空間分解能を持つ状態で計測でき（スペクトルの重心や瞬時周波数を用いる）、同時に、その方向の周波数も計測でき、後の示される周波数の空間積分処理において設定される積分路の方向（接線方向）の周波数を高精度に求めて計算できる。例えば、積分路は、送信時に設定した波動又はビームのステアリング方向（角度）、又は、生成された波動又はビームにおいて、同様に、但し、大局的に推定されるステアリング方向（角度）に直線状に取ることができる。尚、処理を簡単に済ませるべく、公称周波数や同時に大局的に推定されるその方向の周波数を使用することもできる。空間分解能を持つ状態で推定される伝搬方向に積分を行うことは、周波数分布の補間処理を伴い、不可能ではないが、実用的ではない。 Now let the signal at the position coordinate s in the propagation direction (coordinate axis t) in the region of interest obtained using a single wave or beam be

where t=0 represents the reference position in the t-axis direction, that is, the position of the wave source (t=0), and δθ(t) represents the phase change caused by reflection or scattering at the position coordinate t.
Then, from this, the instantaneous angular frequency ω(t) in the propagation direction t, the instantaneous phase θ(t), etc. are obtained and imaged. Propagation direction t can be straight forward without steering, or can be steered to have a deflection angle. Sometimes. As described in Non-Patent Document 19, the propagation direction of waves or beams can be measured with spatial resolution (using the center of gravity of the spectrum and the instantaneous frequency), and at the same time, the frequency in that direction can also be measured. The frequency in the direction (tangential direction) of the integration path set in the spatial integration process can be obtained and calculated with high accuracy. For example, the integration path may be linearly aligned with the steering direction (angle) of the wave or beam set at the time of transmission, or in the generated wave or beam as well, but with a globally estimated steering direction (angle). can take. It should be noted that for ease of processing, the nominal frequency and, at the same time, the globally estimated frequency in that direction can be used. Integrating in the estimated propagation direction with spatial resolution involves interpolating the frequency distribution, which is not impossible, but not practical.

を生成し、式（３０－１）と式（３１）の二乗の和の平方根によっても求まる（特許文献７や非特許文献１４）。後者は、特にデジタル信号処理に適している。 Incidentally, A(s) is the amplitude and represents the reflection intensity or scattering intensity at the position coordinate t=s. For example, it can be obtained by envelope detection (square root of sum of squares of IQ signals) through quadrature detection of equation (30-1). Or, through the Hilbert transform using the Fourier transform,

is also obtained by the square root of the sum of the squares of the equations (30-1) and (31) (Patent Document 7 and Non-Patent Document 14). The latter is particularly suitable for digital signal processing.

と表せる。 Expressed as a complex analytic signal (Patent Document 6 and Non-Patent Document 7) using Equations (30) and (31),

can be expressed as

であると仮定すると、その位置座標ｔ＝ｓ＋Δsにおける信号は、

と表されて、式（３２）と式（３５）との共役積を求めると、式（３３）と式（３４）の仮定の下で、

と表され、従って、位置座標t=ｓにおける瞬時周波数は、

と推定される。 To obtain the instantaneous phase θ(s), first, the instantaneous angular frequency is obtained. As a conventional means, the method described in Patent Document 6 and Non-Patent Document 7 assumes that the instantaneous frequencies at the next sampling position coordinate t = s + Δs are equal and the instantaneous phases are not equal at the position coordinate t = s ( δθ(t) is a phase change (that is, random) determined by random scattering intensity and reflectance, and varies greatly randomly with respect to t),

, the signal at its position coordinate t=s+Δs is

When the conjugate product of equations (32) and (35) is obtained, under the assumptions of equations (33) and (34),

and therefore the instantaneous frequency at the position coordinate t=s is

It is estimated to be.

と、式（３７）そのものに施される場合

とがあるが、変位(ベクトル)計測には、式（３８－１）の方が精度が高いことが確認されている。 As described in Patent Document 6 and Non-Patent Document 7, noise is actually mixed in the signal r(s), and estimation under the assumptions of equations (33) and (34) , s-axis direction and two or one direction orthogonal thereto, moving average processing may be performed to improve accuracy. If this moving average process is applied to equation (36) and determined according to equation (37)

and when applied to equation (37) itself

However, it has been confirmed that formula (38-1) has higher accuracy for displacement (vector) measurement.

を乗算することにより、瞬時位相θ(ｓ)がランダムな散乱強度や反射率で決まる位相変化の積算値（即ち、ランダム）である仮定の下で、その推定値

が得られる。 Using these moving-averaged instantaneous frequencies, the instantaneous frequencies may be detected at each coordinate. The estimate of the instantaneous frequency is unbiased, and for digital signal processing, for equation (32),

By multiplying the estimated value

is obtained.

Instead of the moving average of the instantaneous frequencies obtained by equations (38-1) and (38-2), A frequency (×2π) may be used. The formula is shown in formula (S1).

t=0 in the above equation representing the observed signal represents the reference position in the t-axis direction, that is, the wave source position. On the other hand, the reference position t = t' in equation (39) may also be set to t' = 0 (wave source position). is an estimated value of the instantaneous phase itself (equation (30-2)) represented by the integrated value of the phase change caused by reflection and scattering. The instantaneous frequencies are averaged and the resulting θ'(s) is an estimate determined in that situation.

として式（３９）を用いることとなる。若しくは、t'＝ s'（非零、しかし、ｓとは等しくない）として、式（３９）を用いる場合も有り、その場合には、推定値θ'(s)には、以下のバイアスエラーが生じることになる。

In addition, due to the influence of moving average processing and the influence of the window length when obtaining the spectrum, the instantaneous frequency cannot be obtained from the wave source position (t = 0) to t = s' (nonzero and not equal to s). If not, with t′=0 and with the nominal angular frequency or the previously measured/estimated angular frequency ω ₀ ,

Equation (39) is used as Alternatively, Equation (39) may be used with t'=s' (non-zero but not equal to s), in which case the estimate θ'(s) has a bias error of will occur.

として求まる。式（３６）に対して、ω(s)Δsを核とする複素指数関数との共役積を求めても良い。尚、上記の式（３４）、（３６）、（４３）等において、位相の引き算を前方差分を用いて求める場合を記載したが、その代わりに、後方差分を行うことも可能である。また、式（３７）、式（３８－１）、式（３８－２）において、位相の微分の計算を上記の位相の差分をサンプリング間隔で割る近似で行ったが、高域遮断周波数を持つ微分フィルタを用いて微分処理を行うこともある。また、式（３９）の瞬時周波数の推定値の積分には、台形則を初めとする公知の様々な積分演算を実施できる。
尚、式（４０）で表される位相回転を含まない瞬時位相（式(３０－２)）の推定結果を得るべく、上記に依らずに、式（３２）で表される解析信号の虚数部/実数部にarctan（正接の逆関数）を施して、式（３０－１）中の余弦の角（即ち、位相回転を含む瞬時位相）を求め、式（４２）においてs'=ｓとして瞬時周波数の移動平均又はスペクトルの第１次モーメントの積分演算により求まる位相回転を用いて直接的に減算することもできる。但し、arctanの直接の演算結果は、－π～πの結果となるため、その結果をアンラッピングした上で減算処理を行う必要がある。位相回転を含む瞬時位相は単調増加であるため、アンラッピングは、arctanの結果が負のときに２πを整数ｍ倍したものを足せば良い。但し、ｍはビーム方向又は波動の伝搬方向に観測されたarctanの結果が負となった回数である。尚、上記の場合と同様に、式（４１）を使用する場合もあるし、式（４２）により表されるバイアスエラーを生じることもある。また、バイアスエラーを含まないサンプリング間隔Δｓの瞬時位相の変化量であるΔθ（ｓ）を推定するべく、式（４３）とは別に、Δsだけ隔てた位置にて推定された位相回転を含まない瞬時位相の推定値との差を直接的に引き算により計算することもできる。
式（４０）又は式（４３）を用いて表される位相に関する画像は広帯域化されており、超解像の一種である。それらの位相そのものを表示することもできるし、余弦や正弦関数を掛けて表示することもできるし、さらに、包絡線で重み付けして表示することもできる。ちなみに、式（４０）の位相を求めた複素解析信号の実部と虚部の各々の二乗の和の平方根は、包絡線信号と同一である。従って、二乗検波や絶対値検波、理想的には波打ち（信号値の符号、位相）を壊さずにそのままに画像化すると良い（グレー画像やカラー画像で表示できる）。主として、反射や散乱で決まる信号強度と位相や位相変化を表す画像となる。一方、求められた瞬時周波数を画像化すると、減衰や散乱による周波数変調の影響などを表す画像となる(同じく、グレーやカラー表示できる)。 However, the bias does not matter when estimating Δθ(s), which is the amount of change in the instantaneous phase at sampling intervals Δs, based on equations (30-2) and (34). The estimated result is

is obtained as A conjugate product with a complex exponential function with ω(s)Δs as a kernel may be obtained for Equation (36). In the above equations (34), (36), (43), etc., the phase subtraction is obtained using the forward difference, but it is also possible to use the backward difference instead. In equations (37), (38-1), and (38-2), the phase differentiation is calculated by approximation by dividing the phase difference by the sampling interval. Differential processing may be performed using a differential filter. In addition, various known integration operations such as the trapezoidal rule can be implemented for the integration of the estimated value of the instantaneous frequency in Equation (39).
In order to obtain the estimation result of the instantaneous phase (equation (30-2)) that does not include the phase rotation represented by equation (40), instead of relying on the above, the imaginary number of the analytic signal represented by equation (32) Apply arctan (inverse function of tangent) to the part/real part to obtain the cosine angle (that is, instantaneous phase including phase rotation) in equation (30-1), and set s'=s in equation (42) It is also possible to perform direct subtraction using the moving average of the instantaneous frequency or the phase rotation obtained by integrating the first moment of the spectrum. However, since the direct operation result of arctan is -π to π, the result must be unwrapped before subtraction processing. Since the instantaneous phase including the phase rotation is monotonically increasing, unwrapping can be done by adding 2π multiplied by an integer m when the result of arctan is negative. where m is the number of times the arctan result observed in the beam direction or wave propagation direction is negative. It should be noted that, as in the above case, equation (41) may be used, and the bias error represented by equation (42) may occur. In addition, in order to estimate Δθ(s), which is the amount of change in the instantaneous phase at the sampling interval Δs that does not include the bias error, in addition to Equation (43), the phase rotation estimated at the position separated by Δs is not included. The difference from the instantaneous phase estimate can also be calculated directly by subtraction.
The phase-related image expressed using Equation (40) or Equation (43) is band-widened and is a kind of super-resolution. These phases can be displayed as they are, multiplied by a cosine or sine function, or weighted with an envelope. Incidentally, the square root of the sum of the squares of the real part and the imaginary part of the complex analytic signal whose phase is obtained in Equation (40) is the same as the envelope signal. Therefore, square-law detection, absolute value detection, and ideally imaging without destroying waviness (sign and phase of signal values) (can be displayed as a gray image or a color image). The image mainly represents the signal intensity, phase, and phase change determined by reflection and scattering. On the other hand, when the obtained instantaneous frequency is visualized, it becomes an image that expresses the effects of frequency modulation due to attenuation and scattering (similarly, it can be displayed in gray or color).

それらの信号自身に対して移動平均処理が行われない場合もある。ω(s)は、上記の計算の他、例えば、以下の如く推定することができる。式（３０－１'）をさらに微分し、上記の条件や仮定、処理の下、さらには、ω(s)の空間微分も小さい場合やその仮定をし、以下の様に近似する。

その際に、それらの信号自身に対して移動平均処理は行われない場合もある。式（３０－１''）を式（３０－１）で割り、－１を乗じれば、ω²(s)を推定できる。しかし、ω²(s)のその推定結果が負値となることがあり、偏微分方向sを含む多次元において、又は、その偏微分方向sの近傍の正値に置き換えられたり、近傍の正値のみを使用して補間近似したり、メディアンフィルタが掛けられたり、移動平均処理されたり、これらの組み合わせが施されることがある。メディアンフィルタは、特に、突発的に生じる大きな推定エラーを除去できる効果がある。また、正値に対して平方根を施して得られるω(s)に、メディアンフィルタや移動平均処理が施されることもある。これらの処理を通じて、ω(s)が求められることがある。式（３０－１'）をω(s)で除して－１を乗じることにより、式（３０－１）の解析信号の虚数成分が得られる（つまり、解析信号が求まる）。この２階微分は、式（３０―１）に微分フィルタや差分近似を２回施して求めても良いし、２階の微分フィルタや２階の差分近似（いわゆる中央差分を用いて、１、－２、１を乗じて加算したものをそれらの位置の距離の二乗で除する）を施して得ても良い。この他、微分には、例えば、アナログ回路にてオペアンプを使用した微分フィルタを使用しても良いし、デジタル回路又はデジタル信号処理において、微分フィルタや差分近似に基づく微分計算をしても良い。これらの微分処理は一種の高域通過型フィルタリングであるため、高域遮断周波数を設けて処理したり、微分処理の結果に対して移動平均処理を施すことがある。また、上記の瞬時周波数ω(s)には、計算を簡単にするべく、代わりに、公称周波数や大局的に推定される周波数(スペクトルの第1次モーメント等)を使用することもできる。本検波処理は、他の検波処理よりも格段に高速である。この処理は、r(s)の包絡線イメージング(解析信号の大きさを求めれば良い)や、変位や速度、加速度、歪、歪率、温度等の計測にも使用できる。（高速）フーリエ変換を施した場合と変位の計測精度は略同じであるが、エコー画像は深部において比較して高強度となることを経験している。本ヒルベルト変換法はフーリエ変換を用いたヒルベルト変換（非特許文献１３）よりも高速であり、１つ１つの時相において、波動のパラメータやビームフォーミングパラメータの異なる複数のビームや波動が生成される場合には、受信トランスデューサにて受信する受信信号が増えてビームフォーミングやヒルベルト変換を実施する回数が増えるため、その様な場合に有効である。ビームフォーミングされた複数の信号を重ね合わせたものをビームフォーミングされた１つの信号と見なして一度に本ヒルベルト変換を実施することもあり、その場合にも有効である（微分により求まる瞬時周波数は、それらのビームや波動の重ね合わせの微分方向の合成周波数である）。上記の瞬時位相イメージングも同様に重ね合わされた合成波動において実施されることがある。これらの計測を行う場合には、複数の波動やビームを用いたり、スペクトルの周波数分割等を通じて得られた複数の疑似の波動やビームを用いて、それらの各々から導出されるドプラ方程式を連立して連立方程式を得ることがある（over-determinedシステムであることがある）が、その場合に、式（３０－１）は、それらの各々の波動やビーム、又は、疑似の波動やビームを表し、各々の解析信号が同様にして求められて使用される。また、受信信号が多次元受信信号であり、搬送周波数が複数の直交座標軸方向に存在する場合（横方向変調やステアリングを実施した場合）においても、同様に瞬時位相を推定できる（詳細については後述する）。 Although it has been described that the above Hilbert transform is performed based on the (fast) Fourier transform (Non-Patent Document 13), the original Hilbert transform may of course be calculated. Also, if the signal contains noise, the accuracy will drop, but the real-time signal of interest is subjected to differentiation processing to generate a signal with a 90° phase lead, which is then multiplied by -1 to obtain an imaginary number. It is also possible to calculate the component (the amount multiplied by the angular frequency by differentiation processing is the phase obtained by applying the inverse function of the cosine or sine signal to the original real-time signal, and the same calculation result (phase ) (forward difference, backward difference, central difference, etc., multiplied by 1 and -1 and added or subtracted) and divided by the distance of those positions (difference approximation). (In some cases, differential filtering is applied instead of differential approximation).
For example, when differential processing is applied to an arbitrary signal represented by equation (30-1), the spatial differential (spatial (s) change) of the amplitude A(s) is compared to the instantaneous frequency ω(s). is small or assumed to be small, and for the result of formula (30-1) or its differentiation, by applying a moving average in multiple dimensions including the direction of partial differentiation or in the direction of partial differentiation, the following approximates

Moving average processing may not be performed on those signals themselves. In addition to the above calculation, ω(s) can be estimated, for example, as follows. Equation (30-1') is further differentiated, and under the above conditions, assumptions, and processing, and furthermore, when the spatial differentiation of ω(s) is also small or assumed, the following approximation is obtained.

At that time, moving average processing may not be performed on those signals themselves. ω ² (s) can be estimated by dividing equation (30-1'') by equation (30-1) and multiplying by -1. However, the estimated result of ω ² (s) may be negative, and in multiple dimensions including the partial differential direction s, or replaced by positive values in the neighborhood of the partial differential direction s, or by positive values in the neighborhood Interpolation approximation using only values, median filtering, moving average processing, or a combination of these may be applied. The median filter is particularly effective in removing large estimation errors that occur suddenly. Also, ω(s) obtained by taking the square root of a positive value may be subjected to median filtering or moving average processing. Through these processes, ω(s) may be obtained. By dividing equation (30-1′) by ω(s) and multiplying by −1, the imaginary component of the analytic signal of equation (30-1) is obtained (that is, the analytic signal is obtained). This second derivative may be obtained by applying a differential filter or difference approximation to the equation (30-1) twice, or using a second differential filter or second difference approximation (so-called central difference), 1, −2, divide by the square of the distance between those positions). In addition, for differentiation, for example, a differential filter using an operational amplifier in an analog circuit may be used, or differential calculation based on a differential filter or differential approximation may be performed in a digital circuit or digital signal processing. Since these differentiation processes are a kind of high-pass filtering, they may be processed with a high cut-off frequency or a moving average process may be applied to the results of differentiation processes. Also, for the above instantaneous frequency ω(s), instead, a nominal frequency or a globally estimated frequency (such as the first moment of the spectrum) can be used to simplify the calculations. This detection process is significantly faster than other detection processes. This process can also be used for envelope imaging of r(s) (the magnitude of the analytic signal can be obtained), and measurement of displacement, velocity, acceleration, strain, distortion rate, temperature, and the like. Although the measurement accuracy of the displacement is almost the same as when the (fast) Fourier transform is applied, we have experienced that the echo image has a higher intensity in comparison with the deep part. This Hilbert transform method is faster than the Hilbert transform using the Fourier transform (Non-Patent Document 13), and in each time phase, multiple beams and waves with different wave parameters and beamforming parameters are generated. In some cases, the number of reception signals received by the receiving transducer increases and the number of beamforming and Hilbert transforms increases, so this is effective in such a case. A plurality of beamformed signals superimposed is regarded as one beamformed signal, and this Hilbert transform may be performed at once. is the composite frequency in the differential direction of the superposition of those beams and waves). Instantaneous phase imaging as described above may also be performed on superimposed synthetic waves. When performing these measurements, multiple waves or beams are used, or multiple pseudo waves or beams obtained through frequency division of the spectrum are used, and the Doppler equations derived from each of these are used simultaneously. (which may be an over-determined system), in which case equation (30-1) represents their respective waves or beams, or quasi-waves or beams , each analytic signal is similarly determined and used. In addition, even if the received signal is a multidimensional received signal and the carrier frequencies exist in a plurality of orthogonal coordinate axis directions (when lateral modulation or steering is performed), the instantaneous phase can be similarly estimated (details will be described later). do).

とすると、これより、各方向（t₁,t₂,t₃）の瞬時角周波数（ω₁(t₁,t₂,t₃), ω₂(t₁,t₂,t₃), ω₃(t₁,t₂,t₃)）と、瞬時位相θ(t₁,t₂,t₃)等を求めてイメージングする。関心領域が２次元のときは、

と表される。以下の３次元の時と同様に処理される。 FIG. 35 shows an example of two steering beams in the two-dimensional case and the lateral modulation of their superposition. FIG. 36 also shows an example of four steering beams in the three-dimensional case and the lateral modulation of their superposition (in the case of three steering beams, either It is also possible to use three steering beams out of the three, or alternatively three steering beams that are all symmetrical about the axis may be generated (not shown). Also, although linear one-dimensional and two-dimensional array apertures are shown in FIGS. 35 and 36, other arbitrary shaped array apertures may be used, and the coordinate system may be any Cartesian coordinate system. system can be used. Whichever shape of array aperture or Cartesian coordinate system is used, a steering beam that is spatially symmetrical about the four spatially quadrant planes with axes intersecting the axes (not symmetrical) is used. In some cases, the coordinate system may be translated or rotated to achieve symmetry). Although referred to as a steering beam here, it can be applied to any steered wave (including non-steered ones). In the case of lateral modulation, for example in two dimensions, the two cross waves generated by the steering are processed in non-superimposed and superimposed states, as shown in FIG. In the three-dimensional case, as shown in FIG. 36, three or four crossing waves generated by steering are processed in non-superimposed and superimposed states. The same applies when steering is performed and processed as a multidimensional received signal. As described in Non-Patent Document 19, the propagation direction of waves or beams can be measured with spatial resolution (using the center of gravity of the spectrum and the instantaneous frequency), and at the same time, the frequency in that direction can also be measured. The frequency in the direction (tangential direction) of the integration path set in the spatial integration process can be obtained and calculated with high accuracy. In the case of the above one-dimensional signal, for example, the integration path can be similarly, but globally estimated In this multi-dimensional case, the integration path can theoretically be taken arbitrarily in the dimensional space. However, when actually calculating the integral, it is important to set the integration path in a state suitable for the coordinate system in which beamforming is performed. often used. It is also possible to use the nominal frequency and at the same time a globally estimated frequency (projected onto the integration path) for ease of processing. Integrating in the estimated propagation direction with spatial resolution involves interpolating the frequency distribution, which is not impossible, but not practical. Now let's define the signal in the region of interest as

, the instantaneous angular frequencies _{( ω1 ( t1} ,t2 _, t3), _{ω2 ( t1 , t2 ,} t3), ω ₃ (t ₁ , t ₂ , t ₃ )) and the instantaneous phase θ(t ₁ , t ₂ , t ₃ ), etc. are obtained and imaged. When the region of interest is two-dimensional,

is represented. It is processed in the same way as the 3D case below.

を生成し、式（３０'－１）と式（３１'）の二乗の和の平方根によっても求まる（特許文献７や非特許文献１４）。後者は、特にデジタル信号処理に適している。 Incidentally, A(s ₁ , s ₂ , s ₃ ) is the amplitude and represents the reflection intensity and scattering intensity at the position coordinates (s ₁ , s ₂ , s ₃ ). For example, it can be obtained by envelope detection (square root of sum of squares of IQ signals) through quadrature detection of equation (30'-1). Or, through the Hilbert transform using the Fourier transform,

is also obtained by the square root of the sum of the squares of the equations (30′-1) and (31′) (Patent Document 7 and Non-Patent Document 14). The latter is particularly suitable for digital signal processing.

と表せる。 Using the equations (30′) and (31′), expressed as a complex analytic signal (Patent Document 6 and Non-Patent Document 7),

can be expressed as

であると仮定すると、その位置座標(s₁＋Δs₁,s₂,s₃)における信号は、

と表されて、式（３２'）と式（３５'）との共役積を求めると、式（３３'）と式（３４'）の仮定の下で、

と表され、従って、位置座標(s₁,s₂,s₃)におけるs₁方向の瞬時周波数ω₁(s₁,s₂,s₃)は、

と推定される。 To obtain the instantaneous phase θ(s ₁ , s ₂ , s ₃ ), first, the instantaneous angular frequency is obtained. As a conventional means, at the position coordinates (s ₁ , s ₂ , s ₃ ), for example, the next sampling position coordinates (s ₁ + _{Δs 1 ,} s ₂ , s ₃ ) are assumed to have equal instantaneous frequencies and unequal instantaneous phases (δθ(t ₁ , t ₂ , t ₃ ) are phase changes determined by random scattering intensity and reflectance (i.e., random ), which varies greatly randomly with respect to (t ₁ ,t ₂ ,t ₃ )),

, the signal at its position coordinates (s ₁ +Δs ₁ ,s ₂ ,s ₃ ) is

When the conjugate product of equations (32′) and (35′) is obtained, under the assumptions of equations (33′) and (34′),

Therefore, the instantaneous frequency ω ₁ (s ₁ , s ₂ , s ₃ ) in the s ₁ direction at the position coordinates (s ₁ , s ₂ , s ₃ ) is

It is estimated to be.

及び、ω₁(s₁,s₂,s₃)そのものに施される場合

があるが、変位(ベクトル)計測には、式（３８'―1）の方が精度が高いことが確認されている。ｓ₂やｓ₃の方向の瞬時周波数もＲ₂(s₁,s₂,s₃)やＲ₃(s₁,s₂,s₃)を求めることを通じて同様に求められる。 As described in Patent Document 6 and Non-Patent Document 7, in reality, noise is mixed in the signal r(s ₁ , s ₂ , s ₃ ), and equations (33′) and (34′) As an estimation under the assumption of , moving average processing including the s1 axis direction and two or _one direction orthogonal thereto may be performed to improve accuracy. If this moving average process is applied to equation (36') and determined according to equation (37')

and when applied to ω ₁ (s ₁ , s ₂ , s ₃ ) itself

However, it has been confirmed that formula (38'-1) has higher accuracy for displacement (vector) measurement. The instantaneous frequencies in the directions of s2 and s3 are similarly obtained through _{obtaining R2 (} s1 _, s2 _{, s3} ) and R3 ₍ s1 _, s2 _{, s3} ).

を乗算することにより、瞬時位相θ(s₁,s₂,s₃)がランダムな散乱強度や反射率で決まる位相変化の積算値（即ち、ランダム）である仮定の下で、その推定値

が得られる。 Using these moving-averaged instantaneous frequencies, the instantaneous frequencies may be detected at each coordinate. The instantaneous frequency estimate is unbiased, and for digital signal processing, for equation (32′),

_{By multiplying} the _estimated value

is obtained.

In addition, instead of the moving average of the instantaneous frequencies obtained by the equations (38'-1) and (38'-2), the first moment of the spectrum (the center of gravity, that is, the weighted average value) is The obtained center-of-gravity frequency (×2π) may be used. The formula for the center of gravity in the x-axis direction, one axis of the three-dimensional orthogonal coordinate system, is shown in formula (S1''). Other directions are found in the same way, and also in two dimensions.

The integration path c in the expression representing the above observation signal is an arbitrary line from the starting point ₀ to the interest point ₍ s1, s2 _, s3), where the position where the instantaneous phase is zero is regarded as the reference position. 0 represents the position of the wave source. _On the other hand, the starting point of the integration path _c ' in equation ₍ 39') may also be ₀ (wave source position). θ' (s ₁ , s ₂ , s ₃ ) obtained as the distribution of s ₂ , s ₃ ) is the instantaneous phase (equation (30'-2)) represented by the integral value of the phase change caused by reflection and scattering. is an estimate of The instantaneous frequencies are averaged, and the resulting θ'(s ₁ ,s ₂ ,s ₃ ) is the estimated value obtained in that situation.

として式（３９'）を用いることとなる。若しくは、積分路c'の始点を(t₁,t₂,t₃)＝(s₁',s₂',s₃')（波動源位置０ではなく、(s₁,s₂,s₃)とも等しくない）として、式（３９'）を用いる場合も有り、その場合に、推定値θ'(s₁,s₂,s₃)には、以下のバイアスエラーが生じることになる。

In addition, due to the influence of the moving average process and the influence of the window length when obtaining the spectrum, the instantaneous frequency changes from the wave source position 0 to (t ₁ , t ₂ , t ₃ ) = (s ₁ ', s ₂ ', s ₃ ' ) (not the wave source position and not equal to (s ₁ , s ₂ , s ₃ )), the starting point of the integration path c' is set to 0, and the nominal angular frequency or the previously measured/estimated Using the angular frequencies (ω ₀₁ , ω ₀₂ , ω ₀₃ ) in each direction,

Equation (39′) is used as Alternatively, the starting point of the integration path c' is (t ₁ ,t ₂ ,t ₃ )=(s ₁ ',s ₂ ',s ₃ ') (not the wave source position 0, but (s ₁ ,s ₂ ,s ₃ ) is not equal to )), in which case the following bias error will occur in the estimated value θ'(s ₁ , s ₂ , s ₃ ).

として求まる。式（３６'）に対して、ω₁(s₁,s₂,s₃)Δs₁を核とする複素指数関数との共役積を求めても良い。その他、t₂やt₃の方向の位相変化の推定値Δθ₂'(s₁,s₂,s₃)Δθ₃'(s₁,s₂,s₃)も各方向のサンプリング間隔Δs₂とΔs₃にて求めることができる。尚、上記の式（３４'）、（３６'）、（４３'）等において、位相の引き算を前方差分を用いて求める場合を記載したが、その代わりに、後方差分を行うことも可能である。また、式（３７'）、式（３８'－１）、式（３８'－２）において、位相の微分の計算を上記の位相の差分をサンプリング間隔で割る近似で行ったが、高域遮断周波数を持つ微分フィルタを用いて微分処理を行うこともある。また、式（３９'）の瞬時周波数の推定値の積分には、台形則を初めとする公知の様々な積分演算を実施できる。尚、式（４０'）で表される位相回転を含まない瞬時位相（式(３０'－２)）の推定結果を得るべく、上記に依らずに、式（３２'）で表される解析信号の虚数部/実数部にarctan（正接の逆関数）を施して、式（３０'－１）中の余弦の核（即ち、位相回転を含む瞬時位相）を求め、式（４２'）において(s₁',s₂',s₃')＝(s₁,s₂,s₃)として瞬時周波数の移動平均又はスペクトルの第１次モーメントの積分演算により求まる位相回転を用いて直接的に減算することもできる。但し、arctanの直接の演算結果は、－π～πの結果となるため、その結果をアンラッピングした上で減算処理を行う必要がある。位相回転を含む瞬時位相は単調増加であるため、アンラッピングは、arctanの結果が負のときに2πを整数ｍ倍したものを足せば良い。但し、ｍはビーム方向又は波動の伝搬方向に観測されたarctanの結果が負となった回数である。尚、上記の場合と同様に、式（４１'）を使用する場合もあるし、式（４２'）により表されるバイアスエラーを生じることもある。また、バイアスエラーを含まない、t₁方向のサンプリング間隔Δｓ₁の瞬時位相の変化量であるΔθ₁（s₁,s₂,s₃）を推定するべく、式（４３'）とは別に、t₁方向にΔs₁だけ隔てた位置にて推定された位相回転を含まない瞬時位相の推定値との差を直接的に引き算により計算することもできる。同様にして、t₂やt₃の方向のサンプリング間隔Δs₂とΔs₃の瞬時位相の変化量の推定値Δθ₂'(s₁,s₂,s₃)Δθ₃'(s₁,s₂,s₃)も求めることができる。
式（４０'）又は式（４３'）を用いて表される位相に関する画像は広帯域化されており、超解像の一種である。それらの位相そのものを表示することもできるし、余弦や正弦関数を掛けて表示することもできるし、さらに、包絡線で重み付けして表示することもできる。ちなみに、式（４０'）の位相を求めた複素解析信号の実部と虚部の各々の二乗の和の平方根は、包絡線信号と同一である。従って、二乗検波や絶対値検波、理想的には波打ち（信号値の符号、位相）を壊さずにそのままに画像化すると良い（グレー画像やカラー画像で表示できる）。 However, for example, based on equations (30′-2) and (34′), Δθ ₁ (s ₁ , s ₂ , s ₃ ), which is the amount of change in the instantaneous phase in the t ₁ -axis direction at sampling intervals Δs ₁ , is When estimating, the bias does not matter. The estimated result is

is obtained as A conjugate product with a complex exponential function having ω ₁ (s ₁ , s ₂ , s ₃ )Δs ₁ as a kernel may be obtained for equation (36′). In addition, the estimated values Δθ ₂ '(s ₁ , s ₂ , s ₃ ) Δθ ₃ '(s ₁ , s ₂ , s ₃ ) of the phase change in the directions of t ₂ and t ₃ are also the sampling intervals Δs ₂ and It can be obtained by _Δs3 . In the above equations (34′), (36′), (43′), etc., the case of obtaining the phase subtraction using the forward difference has been described, but it is also possible to perform the backward difference instead. be. In equations (37'), (38'-1), and (38'-2), the phase differentiation is calculated by approximation by dividing the phase difference by the sampling interval. Differential processing may be performed using a differential filter having a frequency. In addition, various known integration operations such as the trapezoidal rule can be implemented for the integration of the estimated value of the instantaneous frequency of Equation (39'). In order to obtain the estimation result of the instantaneous phase (equation (30'-2)) that does not include the phase rotation represented by equation (40'), the analysis represented by equation (32') is performed without depending on the above. Apply arctan (inverse tangent function) to the imaginary part/real part of the signal to obtain the cosine kernel (that is, the instantaneous phase including the phase rotation) in equation (30'-1), and in equation (42') (s ₁ ', s ₂ ', s ₃ ') = (s ₁ , s ₂ , s ₃ ) directly using the phase rotation obtained by the moving average of the instantaneous frequency or the integral operation of the first moment of the spectrum You can also subtract. However, since the direct operation result of arctan is -π to π, the result must be unwrapped before subtraction processing. Since the instantaneous phase including the phase rotation is monotonically increasing, unwrapping can be done by adding 2π multiplied by an integer m when the result of arctan is negative. where m is the number of negative arctan results observed in the beam direction or wave propagation direction. It should be noted that, as in the above case, the equation (41') may be used, and the bias error represented by the equation (42') may occur. In addition, in order to estimate Δθ ₁ (s ₁ , s ₂ , s ₃ ), which is the amount of change in the instantaneous phase at the sampling interval Δs ₁ in the t ₁ direction and does not include the bias error, in addition to Equation (43′), It is also possible to calculate the difference from the estimated value of the instantaneous phase not including the phase rotation estimated at the position separated by Δs ₁ in the t ₁ direction by direct subtraction. Similarly, the estimated values Δθ ₂ '(s ₁ ,s _{2 ,s 3 ) Δθ 3} ' ₍ s ₁ , _{s 2} ,s ₃ ) can also be obtained.
The phase-related image expressed using Equation (40′) or Equation (43′) is band-widened and is a kind of super-resolution. These phases can be displayed as they are, multiplied by a cosine or sine function, or weighted with an envelope. Incidentally, the square root of the sum of the squares of the real part and the imaginary part of the complex analytic signal for which the phase of Equation (40') is obtained is the same as the envelope signal. Therefore, square-law detection, absolute value detection, and ideally imaging without destroying waviness (sign and phase of signal values) (can be displayed as a gray image or a color image).

それらの信号自身に対して移動平均処理が行われない場合もある。例えば、横方向変調の場合には、方向に依ってはその方向の瞬時周波数が他の方向の瞬時周波数よりも低いことが有り、その様な場合には、偏微分の方向をその方向に取ることも可能であるが、ステアリングした方向であるとすると、信号のサンプリング間隔が粗いことが有り、近似計算を行うに当たり注意を要する。ω₁(s₁,s₂,s₃)は、上記の計算の他、例えば、以下の如く推定することができる。式（３０'－１'）を同一方向にさらに偏微分し、上記の条件や仮定、処理の下、さらには、ω₁(s₁,s₂,s₃)の空間微分も小さい場合やその仮定をし、以下の様に近似する。

その際に、それらの信号自身に対して移動平均処理は行われない場合もある。式（３０'－１''）を式（３０'－１）で割り、-1を乗じれば、ω₁ ²(s₁,s₂,s₃)を推定できる。しかし、ω₁ ²(s₁,s₂,s₃)のその推定結果が負値となることがあり、偏微分方向を含む多次元において、又は、その偏微分方向s₁の近傍の正値に置き換えられたり、近傍の正値のみを使用して補間近似したり、メディアンフィルタが掛けられたり、移動平均処理されたり、これらの組み合わせが施されることがある。メディアンフィルタは、特に、突発的に生じる大きな推定エラーを除去できる効果がある。また、正値に対して平方根を施して得られるω₁(s₁,s₂,s₃)に、メディアンフィルタや移動平均処理が施されることもある。これらの処理を通じて、ω₁(s₁,s₂,s₃)が求められることがある。式（３０'－１'）をω₁(s₁,s₂,s₃)で除して-1を乗じることにより、式（３０'－１）の解析信号の虚数成分が得られる（つまり、解析信号が求まる）。この２階微分は、式（３０'―１）に微分フィルタや差分近似を２回施して求めても良いし、２階の微分フィルタや２階の差分近似（いわゆる中央差分）を施して得ても良い。s₁以外のs₂やs₃の各々の方向に偏微分した場合も、同様にして、ω₂(s₁,s₂,s₃)とω₃(s₁,s₂,s₃)が求まり、解析信号が求まるが、上記の通り、偏微分の方向を適切に選ぶことが望ましい。例えば、２次元の場合において、±２０°のステアリング角度のステアリングビームを用いた場合では、偏微分の処理の方向を深さ方向とした場合と横方向にした場合との結果に大きな違いが無いケースを経験しているが、画像化において、多次元(高速)フーリエ変換を通じたヒルベルト変換(非特許文献１３)と比較すると異なり、本近似処理を用いると比較して深部のエコー強度が強く求まることがある。一方、その様な場合でも、変位ベクトル計測の精度は殆どに同じであることを経験している。この他、微分には、例えば、アナログ回路にてオペアンプを使用した微分フィルタを使用しても良いし、デジタル回路又はデジタル信号処理において、微分フィルタや差分近似に基づく微分計算をしても良い。これらの微分処理は一種の高域通過型フィルタリングであるため、高域遮断周波数を設けて処理したり、微分処理の結果に対して移動平均処理を施すことがある。また、上記のω₁(s₁,s₂,s₃)等の瞬時周波数には、計算を簡単にするべく、代わりに、公称周波数や大局的に推定される周波数(スペクトルの第1次モーメント等)を使用することもできる。本検波処理は、他の検波処理よりも格段に高速である。この処理は、r(s₁,s₂,s₃)の包絡線イメージング(解析信号の大きさを求めれば良い)や、変位(ベクトル)や速度(ベクトル)、加速度(ベクトル)、歪(テンソル)、歪率(テンソル)等の計測にも使用できる。ベクトルやテンソルの計測を行う場合には、複数の波動やビームを用いたり、スペクトルの周波数分割等を通じて得られた複数の疑似の波動やビームを用い、それらの各々から導出されるドプラ方程式を連立して連立方程式を得る（横方向変調やover-determinedシステムであることがある）が、その場合に、式（３０'－１）は、それらの各々の波動やビーム、疑似の波動やビームを表し、各々の解析信号が同様にして求められて使用される。その他、同様にして、温度等の計測に使用されることもある。
本ヒルベルト変換法は多次元フーリエ変換を用いたヒルベルト変換（非特許文献１３）よりも高速であり、１つ１つの時相において、波動のパラメータやビームフォーミングパラメータの異なる複数のビームや波動が生成される場合（但し、ステアリング方向が同一であったりするし、ステアリング方向が軸に対して対称であるとも限らない）には、受信トランスデューサにて受信する受信信号が増えてビームフォーミングやヒルベルト変換を実施する回数が増えるため、その様な場合に有効である。ビームフォーミングされた複数の信号を重ね合わせたものをビームフォーミングされた１つの信号（式（３０'－１））と見なして一度に本ヒルベルト変換を実施することもあり、その場合にも有効である（偏微分により求まる瞬時周波数は、それらのビームや波動の重ね合わせの偏微分方向の合成周波数である）。上記の瞬時位相イメージングも同様に重ね合わされた合成波動において実施されることがある。物理開口素子が２次元又は３次元分布や多次元のアレイを構成している場合において、多くの処理時間を要するという問題をさらに効果的に解決し、本ヒルベルト変換の高速性は有効となる。 In addition, the above Hilbert transform should be performed based on the multidimensional (fast) Fourier transform (Non-Patent Document 13, even if the cross beams and cross waves overlap in space, they are separated without overlapping. can also be processed, but as will be confirmed later, the former process can perform the Fourier transform of all received signals at once, requiring less computation, and the separated signals are actively superimposed before being processed. However, the original Hilbert transform calculation may be applied as in the one-dimensional case.
In addition, although the accuracy decreases when noise is mixed in the signal, a signal whose phase is advanced by 90° is generated by performing differentiation processing on the real-time signal of interest, as in the one-dimensional case. can also be multiplied by -1 to calculate the imaginary component (the amount multiplied by the angular frequency through differentiation processing is the phase obtained by multiplying the original real-time signal by the inverse function of the cosine or sine signal). The difference (so-called forward difference, backward difference, central difference, etc.) from the same calculation result (phase) at the position of is calculated and corrected by dividing it by the distance of those positions (difference approximation).In difference approximation, may be applied with a differential filter). However, the method described in this paragraph is limited to cases where the steered waves are not superimposed, as shown in FIGS. 35 and 36, when lateral modulation is being performed. It is also possible to process a single steered wave alone (this paragraph). One processing method for overlapping cases is described at the end of this paragraph, and the other two processing methods are described in the following paragraphs.
For example, when performing partial differentiation processing on an arbitrary signal represented by equation (30'-1) or the like, partial differentiation is performed _{on one} direction from s1 to s3. _The spatial partial derivative (spatial change) of the amplitude _{A (} s1,s2 _, s3) is the instantaneous frequency ω1 ₍ s1,s2 _{, s3} ) or ω2 _{( s1} ,s ₂ , s ₃ ) and ω ₃ (s ₁ , s ₂ , s ₃ ), and ω 3 (s 1 , s 2 , s 3 ), or assuming the same. The following approximation is obtained by applying a moving average in multiple dimensions including or in the direction of partial differentiation. For example, partial differentiation in the _s1 direction is as follows.

Moving average processing may not be performed on those signals themselves. For example, in the case of transverse modulation, depending on the direction, the instantaneous frequency in that direction may be lower than the instantaneous frequency in other directions, in which case the direction of the partial derivative is taken in that direction. However, if it is in the steered direction, the sampling interval of the signal may be rough, and care must be taken in performing approximate calculations. ω ₁ (s ₁ , s ₂ , s ₃ ) can be estimated as follows, for example, in addition to the above calculation. (30'-1') is further partially differentiated in the same direction, and under the above conditions, assumptions, and processing, furthermore, when the spatial derivative of ω ₁ (s ₁ , s ₂ , s ₃ ) is also small, or We make an assumption and approximate it as follows.

At that time, moving average processing may not be performed on those signals themselves. ω ₁ ² (s ₁ , s ₂ , s ₃ ) can be estimated by dividing equation (30'-1'') by equation (30'-1) and multiplying by -1. However, the estimated result of ω ₁ ² (s ₁ , s ₂ , s ₃ ) may be negative, and positive values in multiple dimensions including the partial differential direction or near the partial differential direction s ₁ , interpolation approximation using only nearby positive values, median filtering, moving average processing, or a combination of these. The median filter is particularly effective in removing large estimation errors that occur suddenly. Also, ω ₁ (s ₁ , s ₂ , s ₃ ) obtained by square rooting positive values may be subjected to median filtering or moving average processing. ω ₁ (s ₁ , s ₂ , s ₃ ) may be obtained through these processes. Dividing equation (30′-1′) by ω ₁ (s ₁ ,s ₂ ,s ₃ ) and multiplying by −1 gives the imaginary component of the analytic signal of equation (30′-1) (that is, , the analytic signal is obtained). This second derivative may be obtained by applying a differential filter or difference approximation to Equation (30′-1) twice, or by applying a second differential filter or second order difference approximation (so-called central difference). can be Similarly, when partial differentiation is performed in _each direction of s2 and s3 _{other than} s1, ω2 ₍ s1, s2 _{, s3} ) and ω3 ₍ s1 _, s2 _{, s3} ) _are Although the analytic signal is obtained, it is desirable to appropriately select the direction of partial differentiation as described above. For example, in the two-dimensional case, when a steering beam with a steering angle of ±20° is used, there is no significant difference in the results between the depth direction and the horizontal direction of partial differentiation processing. I have experienced a case, but in imaging, unlike Hilbert transform (Non-Patent Document 13) through multidimensional (fast) Fourier transform, deep echo intensity can be obtained strongly compared to using this approximation process. Sometimes. On the other hand, we have experienced that the accuracy of the displacement vector measurement is almost the same even in such a case. In addition, for differentiation, for example, a differential filter using an operational amplifier in an analog circuit may be used, or differential calculation based on a differential filter or differential approximation may be performed in a digital circuit or digital signal processing. Since these differentiation processes are a kind of high-pass filtering, they may be processed with a high cut-off frequency or a moving average process may be applied to the results of differentiation processes. Also, for the instantaneous frequencies such as ω ₁ (s ₁ , s ₂ , s ₃ ) above, instead of the nominal frequency or the globally estimated frequency (first moment of the spectrum etc.) can also be used. This detection process is significantly faster than other detection processes. This processing includes envelope imaging of r(s ₁ , s ₂ , s ₃ ) (the magnitude of the analytic signal can be obtained), displacement (vector), velocity (vector), acceleration (vector), strain (tensor ), distortion factor (tensor), etc. When measuring vectors and tensors, multiple waves and beams are used, and multiple pseudo waves and beams obtained through spectral frequency division are used, and Doppler equations derived from each of them are simultaneously generated. to obtain a system of equations (which may be a transversely modulated or over-determined system), where Eq. , and each analytic signal is similarly determined and used. In addition, it may also be used to measure temperature and the like in the same way.
This Hilbert transform method is faster than the Hilbert transform using the multidimensional Fourier transform (Non-Patent Document 13), and in each time phase, multiple beams and waves with different wave parameters and beamforming parameters are generated. (However, the steering direction may be the same, and the steering direction may not be symmetrical with respect to the axis), the received signal received by the receiving transducer increases and beamforming and Hilbert transform are performed. It is effective in such a case because the number of times of execution increases. A plurality of beamformed signals superimposed may be regarded as one beamformed signal (equation (30′-1)) and the Hilbert transform may be performed at once. (The instantaneous frequency obtained by partial differentiation is the synthesized frequency in the direction of partial differentiation of the superposition of those beams and waves). Instantaneous phase imaging as described above may also be performed on superimposed synthetic waves. In the case where the physical aperture elements form a two-dimensional or three-dimensional distribution or a multi-dimensional array, the problem of requiring a long processing time can be effectively solved, and the high-speed Hilbert transform is effective.

と表される場合に、s₁に関する１階偏微分及び２階偏微分を求めると、

であり、(３０A－１)及び(３０A－３)より、上記と同様に、瞬時角周波数ω₁(s₁,s₂)が求まり、(３０A－２)より、

が求まる。ここで、（３０A－１）では、s₁とs₂の方向の余弦関数の初期値は省略している。また、(３０A－２)をs₂に関して偏微分すると、

であり、（３０A－１）をs₂に関する１階偏微分及び２階偏微分を求めて同様にして求まる瞬時角周波数ω₂(s₁,s₂)と、上記にて求めた瞬時角周波数ω₁(s₁,s₂)とを用いて、

が求まる。(３０A－６)を求めるためには、(３０A－１)をs₂の次にs₁で偏微分しても良く、その場合には、（３０A－４）の代わりに、

も求まる。(３０A－１)、(３０A－４)、(３０A－６)、(３０A－７)の加算や減算により、解析信号

又は、

と、

又は、

とを求め、それらの２つの独立な解析信号に多次元自己相関法や多次元ドプラ法（非特許文献１３）、デモジュレーション法（特許文献７）等の変位計測法を施して変位ベクトルを求めることができ、また、各々の包絡線検波も得られ、それらの各々が画像化されることがある。複数求まる包絡線検波の結果は重ね合わせて画像化に使用しても良い(スペックルの低減効果がある)。検波は包絡線検波に限らず、二乗検波等の他の検波処理を施すことが可能であり、解析信号の実数成分及び/又は虚数成分を使用でき、それらの各々を画像化に使用できるし、複数求まる検波の結果は重ね合わせて画像化に使用しても良い(スペックルの低減効果がある)。
実際の(３０A－６)の計算結果は、偏微分の順番を逆にすると異なるため、(３０A－６)をその様にして２つ求めて、(３０A－８)、(３０A－８')、(３０A－９)、(３０A－９')中にて使用し、同一の式で表される解析信号は重ね合わせて平均化して使用しても良い(変位計測や画像化)。また、未知変位ベクトル成分に関するover-determinedシステムを実現しても良い（変位計測や画像化）。複数求まる包絡線検波の結果は重ね合わせて画像化に使用しても良い（スペックルの低減効果がある）。
また、交差ビームを重ねて横方向変調の施されたエコー信号そのものは、以下の如くに、

の何れかとして表されることもあり、(３０A－１)に施した同じ手順で解析信号やover-determinedシステムを求めて用いることができる。尚、(３０A－１')、(３０A－１'') 、(３０A－１''') においては、(３０A－１)と同様に、s₁とs₂の方向の余弦関数や正弦関数の初期値は省略している。
また、３次元の場合において、３つ又は４つの交差波動を重ねて横方向変調の施されたエコー信号そのものが、

と表される場合に、２次元の場合と同様に偏微分処理を通じて、解析信号

と、これと独立な解析信号を合せ、交差波動の総数（３つ又は４つ）だけ求めることができ、その内の少なくとも３つを連立するか、未知変位成分は３つであるため全てを連立してover-determinedシステムが得られる。ここで、(３０B－１)においてs₁とs₂とs₃の方向の余弦関数の初期値は省略している。２次元の場合と同様に、３次元エコーイメージングも可能である。包絡線検波や二乗検波等の検波を通じて各々の信号が画像化されることがある。また、重ね合わせて画像化されることもある（スペックルの低減効果もある）。複数の検波信号を２次元の場合と同様に偏微分方向の順番を変えて同一の式で表される解析信号を複数個求めることもでき、それらを用いて、over-determinedシステムを実現して、変位計測や画像化が行われることもある（スペックルの低減効果もある）。
また、交差ビームを重ねて横方向変調の施されたエコー信号そのものは、以下の如くに、

の何れかで表されることもあり、同様に、３つ又は４つの解析信号、偏微分処理を施す方向の順番を変えた場合の信号を求め、使用することができる。但し、(３０B－１')、(３０B－１'') 、(３０B－１''')、(３０B－１'''')、(３０B－１''''')、(３０B－１'''''')、(３０B－１''''''')においては、(３０B－１)と同様に、s₁とs₂とs₃の方向の余弦関数や正弦関数の初期値は省略している。 Further, the echo signal itself, which has undergone lateral modulation by superimposing crossing beams or crossing waves, can be directly processed as follows (Fig. 35 for two dimensions, Fig. 35 for three dimensions) 36).
In the two-dimensional case, the echo signal itself, which has been laterally modulated by overlapping two cross waves, can be directly processed as follows. The echo signal is

When obtaining the first and second partial derivatives with respect to s ₁ ,

From (30A-1) and (30A-3), the instantaneous angular frequency ω ₁ (s ₁ , s ₂ ) is obtained in the same manner as above, and from (30A-2),

is sought. Here, in (30A- ₁ ), the initial values of the cosine functions in the directions of s1 and s2 _are omitted. Also, when (30A- ₂ ) is partially differentiated with respect to s2,

The instantaneous angular frequency ω ₂ (s ₁ , s ₂ ) obtained in the same manner by obtaining the first and second partial differentials of (30A-1) with respect to s ₂ and the instantaneous angular frequency obtained above Using ω ₁ (s ₁ , s ₂ ) and

is sought. To obtain (30A-6), (30A _{- 1} ) may be partially differentiated by s2 and then by s1, in which case instead of (30A-4),

is also sought. The analytic signal is

or

When,

or

are obtained, and displacement measurement methods such as the multidimensional autocorrelation method, the multidimensional Doppler method (Non-Patent Document 13), and the demodulation method (Patent Document 7) are applied to these two independent analytic signals to obtain the displacement vector. can also be obtained and each of them can be imaged. A plurality of envelope detection results obtained may be superimposed and used for imaging (there is a speckle reduction effect). Detection is not limited to envelope detection, other detection processing such as square law detection can be applied, real and/or imaginary components of the analytic signal can be used, each of which can be used for imaging, A plurality of detection results obtained may be superimposed and used for imaging (there is a speckle reduction effect).
Since the actual calculation result of (30A-6) is different if the order of partial differentiation is reversed, two (30A-6) are obtained in that way, (30A-8), (30A-8') , (30A-9), and (30A-9'), and the analytic signals expressed by the same formula may be superimposed and averaged for use (displacement measurement and imaging). Also, an over-determined system for unknown displacement vector components may be implemented (displacement measurement and imaging). A plurality of envelope detection results obtained may be superimposed and used for imaging (there is a speckle reduction effect).
Also, the echo signal itself, which has been laterally modulated by superimposing the crossing beams, is as follows:

and can be used to obtain the analytic signal and the over-determined system by the same procedure applied to (30A-1). In (30A-1'), (30A-1'') and (30A _{- 1} '''), the cosine function and sine function The initial value of is omitted.
Also, in the case of three dimensions, the echo signal itself, which has been laterally modulated by superimposing three or four crossing waves, is

, the analytic signal

and independent analytic signals can be combined to obtain the total number of crossing waves (three or four). Together, an over-determined system is obtained. Here, the initial values of the cosine functions in the directions of s ₁ , s ₂ and s ₃ are omitted in (30B-1). As in the two-dimensional case, three-dimensional echo imaging is also possible. Each signal may be imaged through detection such as envelope detection or square law detection. In addition, they may be superimposed and imaged (there is also the effect of reducing speckle). A plurality of analytic signals expressed by the same equation can be obtained by changing the order of the partial differential direction of a plurality of detected signals in the same manner as in the two-dimensional case. , displacement measurement and imaging may be performed (there is also a speckle reduction effect).
Also, the echo signal itself, which has been laterally modulated by superimposing the crossing beams, is as follows:

Similarly, it is possible to obtain and use three or four analytic signals and signals obtained by changing the order of the directions in which partial differential processing is performed. However, (30B-1'), (30B-1''), (30B-1'''), (30B-1''''), (30B-1'''''), (30B- 1'''''') and (30B-1''''''), similar to (30B _{- 1} ), the cosine and sine functions in the directions of s1 _, s2 and s3 Initial values are omitted.

及び、

を得て、さらに、(３０C－１)又は(３０C－２)に対し、虚数項のcos関数の座標（即ち、s₂とs₁）に関して同様に処理を施し、

を得る。これらの(３０C－１)、(３０C－２)、(３０C－３)の実数項(それらにおいて、共通)と虚数項より、解析信号(３０A－８)、(３０A－８')、(３０A－９)、(３０A－９')を求めることができ、上記の如くこれらの４式の内の独立な２つを求めて、多次元自己相関法や多次元ドプラ法（非特許文献１３）、デモジュレーション法（特許文献７）を用いて、変位ベクトルを求めることができる。また、それらの独立した２式の内の各々の包絡線検波や二乗検波等の検波の結果も得られる（スペックルの低減効果もある）。計算量に関して、トータルの１次元フーリエ変換又は１次元逆フーリエ変換の回数は、非特許文献１３にて報告した２次元フーリエ変換を実施する方法の回数と同じである（計６方向の回数）。
また、３次元の場合に、３つ又は４つの交差波が重なった横方向変調のエコー信号が、(３０B－１)と表わされるとすると、２次元の場合と同様に、１次元フーリエ変換と１次元逆フーリエ変換を通じて、解析信号(３０B－２)とこれと独立な解析信号を合せて交差波動の総数（３つ又は４つ）だけ求めることができ、その内の少なくとも３つを連立するか、未知変位成分は３つであるため全てを連立してover-determinedシステムが得られる。２次元の場合と同様に、３次元エコーイメージングも可能である。非特許文献１３にて報告した３次元フーリエ変換を用いる方法と同一回数の１次元フーリエ変換と１次元逆フーリエ変換を実施することになる(交差波が３つのときは計１２方向の回数、交差波が４つのときは計１５方向の回数)。
これらの場合においても、同様に変位ベクトルの計測が行われる。また、包絡線検波や二乗検波等の検波を通じて各々の信号が画像化されることがある。また、重ね合わせて画像化されることもある（スペックルの低減効果もある）。over-determinedシステムである場合にも、同様に変位計測や画像化が行われることがある（スペックルの低減効果もある）。
以上、交差波が重なって横方向変調が実施された信号そのものに関する新しい高速なヒルベルト変換法を記載したが、その重なった交差波とは、同時に波動を送信した場合や各々の送信に対する受信信号を重ねたものである。これに対し、その前に記載した交差波が分離されている場合（即ち、各々が送信された際の受信信号や同時に送信された際の受信信号が処理されて分離された信号）には、例えば、２次元の場合に２つの交差波が重なっていないときに独立な２つの波動信号（(３０A－８)又は(３０A－８')と(３０A－９)又は(３０A－９')の実信号）が、

及び、

として表される場合には、各々に１次元フーリエ変換と１次元逆フーリエ変換を2方向の回数ずつ（即ち、２次元フーリエ変換と２次元逆フーリエ変換を1回ずつ）施すこととなり、計８方向の回数の変換処理が必要となる。非特許文献１３や上記の通りに、分離されている受信信号は重ね合わせし、また、同時に複数の波動を受信した場合の方が計算量が少ない。３次元の場合も同様である(１次元変換処理を、交差波が３つのときは計１８方向の回数、交差波が４つのときは計２４方向の回数；３次元変換処理としては、各々、６回と８回)。 In addition, in the lateral modulation (FIG. 35 for two dimensions, FIG. 36 for three dimensions) by superposition of crossing waves, Hilbert transform using multidimensional (fast) Fourier transform (Non-Patent Document 13) , and the partial differential processing in the previous paragraph, it is also possible to implement the Hilbert transform by applying a one-dimensional (fast) Fourier transform.
In the two-dimensional case, if a two-dimensional echo signal with superimposed crossing waves is expressed as (30A-1), a one-dimensional (fast) Fourier transform is performed on _each of s1 and s2, and _each halfband is After zero-filling the spectrum of (negative band), one-dimensional inverse Fourier transform is performed,

as well as,

Further, (30C-1) or (30C-2) is similarly processed with respect to the cosine function coordinates of the imaginary term (that is, s ₂ and s ₁ ),

get From these (30C-1), (30C-2), (30C-3) real terms (common among them) and imaginary terms, the analytic signals (30A-8), (30A-8'), (30A -9), (30A-9') can be obtained, and as described above, independent two of these four equations are obtained, and the multidimensional autocorrelation method and the multidimensional Doppler method (Non-Patent Document 13) , the displacement vector can be determined using the demodulation method (Patent Document 7). In addition, the results of detection such as envelope detection and square-law detection for each of these two independent equations can also be obtained (there is also a speckle reduction effect). In terms of computational complexity, the number of total one-dimensional Fourier transforms or one-dimensional inverse Fourier transforms is the same as the number of methods for performing two-dimensional Fourier transforms reported in Non-Patent Document 13 (6 directions in total).
Also, in the case of three dimensions, if the laterally modulated echo signal in which three or four intersecting waves are superimposed is represented by (30B-1), one-dimensional Fourier transform and Through one-dimensional inverse Fourier transform, the total number of crossing waves (3 or 4) can be obtained by combining the analytic signal (30B-2) and the independent analytic signal, and at least three of them can be simultaneously Or, since there are three unknown displacement components, an over-determined system can be obtained by combining all of them. As in the two-dimensional case, three-dimensional echo imaging is also possible. One-dimensional Fourier transform and one-dimensional inverse Fourier transform are performed the same number of times as the method using the three-dimensional Fourier transform reported in Non-Patent Document 13 (when there are three crossing waves, a total of 12 When there are 4 waves, the number of times in a total of 15 directions).
In these cases, displacement vectors are similarly measured. Further, each signal may be imaged through detection such as envelope detection and square-law detection. In addition, they may be superimposed and imaged (there is also the effect of reducing speckle). In the case of an over-determined system, displacement measurement and imaging may also be performed in the same way (with the effect of reducing speckle).
So far, we have described a new high-speed Hilbert transform method for the signal itself in which cross-waves are superimposed and laterally modulated. It is layered. On the other hand, if the previously described crossing waves are separated (i.e., the signals received when each was transmitted or the received signals when transmitted simultaneously were processed and separated), For example, two independent wave signals ((30A-8) or (30A-8') and (30A-9) or (30A-9') when two crossing waves do not overlap in the two-dimensional case) real signal) is

as well as,

, each is subjected to a one-dimensional Fourier transform and a one-dimensional inverse Fourier transform twice in two directions (that is, a two-dimensional Fourier transform and a two-dimensional inverse Fourier transform once each), for a total of 8 Direction conversion processing is required for the number of times. As described in Non-Patent Document 13 and above, the amount of calculation is smaller when the separated received signals are superimposed and when a plurality of waves are received at the same time. The same applies to the three-dimensional case (one-dimensional conversion processing is performed a total of 18 times in 18 directions when there are 3 crossing waves, and a total of 24 directions when there are 4 crossing waves; 6 and 8 times).

When this method is used, the image mainly represents the signal intensity, phase, and phase change determined by reflection and scattering. On the other hand, when the obtained instantaneous frequency is imaged, it becomes an image that shows the effects of frequency modulation due to attenuation and scattering (similarly, it can be displayed in gray or color). It should be noted that the existence of the above instantaneous phase deteriorates the measurement accuracy of any of the above tissue displacement measurement methods based on the Doppler method or the classical measurement method, even if the displacement is very small. However, the inventors of the present application have overcome this by developing a phase matching method between frames (for example, Non-Patent Document 15). Other reported methods of expanding and contracting signals representing tissue deformation are also effective, but the former phase matching method is not suitable for high-intensity and random signals in tissue displacement and strain measurement. Essential for measurement (translation and rotation are possible). Normally, blood flow is measured using a narrow band signal, but this method enables high-resolution measurement and opens the way to high-precision viscosity measurement. It is also possible to measure multidimensional vectors and tensors. Blood flow measurement can also be subjected to a detailed examination by performing phase matching processing.

The above envelope detection is conventionally performed on the generated image signal. Their operation in terms of wavenumber (frequency) components is also useful (one of the nonlinear processes in the present invention). In addition to the above, square-law detection, absolute value detection, or the like may be applied for amplitude detection. It should be noted that what is obtained through Fourier transform from an image signal obtained by beamforming (that is, focus and deflection achieved by applying delay and apodization) is a spectrum. If so, the axial and at least one lateral Fourier transform is the angular spectrum. That is, beamforming processing may be further performed after the image signal is generated. In addition, super-resolution processing including this processing and other processing (spectrum weighting, nonlinear processing of signals, inverse filtering, etc., which will be described in detail later) is used for the above coherent addition and may be used for the incoherent addition described above. Objects of addition are different or same signals (before or after beamforming) subjected to other processing, raw signals thereof, or the like. Coherent addition is suitable for widening the bandwidth (high resolution) and increasing the SN ratio, while incoherent addition is suitable for reducing speckle and increasing the SN ratio. The reduction of speckle may be accompanied by a decrease in spatial resolution, but processing with super-resolution does not pose a problem and may bring about a result of high spatial resolution. Incoherent addition is basically applied to the values that have been subjected to some kind of detection (including power detection) and converted to positive values. It is a process that leaves coherence in the signal (at least you can see the wave oscillations). Envelope detection is also useful, but detection with such coherence is particularly useful as detection processing without impairing spatial resolution. In comparison, envelope detection tends to reduce the spatial resolution.

The mode of operation may be determined by commands (signals) input to the device. In addition, waves to be observed (types and characteristics of waves, intensity, frequency, bandwidth, or sign, etc.), propagating media (propagation velocity, physical properties related to waves, attenuation, scattering, transmission, reflection, refraction, or , their frequency dispersion, etc.) may be provided, and the received signals may be analog or digitally processed as appropriate. The characteristics of the generated image signal (such as intensity, frequency, bandwidth or sign) may also be analyzed. The data obtained by the device according to this embodiment may be used in other devices. The apparatus according to the present embodiment may be used as one of network devices, may be controlled by a network system control device, or may be used as a control device for controlling network devices. , may also be a controller that controls a locally configured network.

When the passive device according to this embodiment is used as an active device, the transmission transducer (or applicator) 10 is connected to the transmission unit 31 of the device main body 30, and the transmitter 31a is an analog device, and the trigger signal , the trigger signal is generated and applied by the control unit 34, or if there is a mode in which the transmitter 31a is a digital device and operates according to an external clock signal, either in the device main body 30 or from the control unit 34, or the device body 30 operates with the clock signal of the transmitter 31a. If the transmitter 31a is a digital device, one of these mechanisms synchronizes the transmit and receive clock signals. This is important when generating an image signal based on multiple transmissions. In the absence of synchronization, errors may be reduced in situations where the clock frequency or sampling frequency is high.

As described above, arbitrary beamforming can be realized by digital processing through fast Fourier transform without performing interpolation approximation at high speed. Virtually arbitrary focusing and arbitrary steering (deflection) can be performed with transducer array devices of arbitrary aperture shape. In the present invention, as described in methods (1) to (7), interpolation approximation processing may be performed in arbitrary beamforming wavenumber matching, and beamforming may be performed at a higher speed. In order to perform approximate wavenumber matching with high accuracy, it is necessary to appropriately oversample the received signal at the expense of an increase in the amount of calculation. In that case, it should be noted that the number of data for Fourier transform increases, unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed. The second embodiment can be used in conventional devices as a device and in terms of modes of operation (e.g., imaging mode, Doppler mode, measurement mode, communication mode, etc.), and those and those described above. is not limited to

In the above first and second embodiments, electromagnetic waves, vibration waves (mechanical waves) including sound waves (compression waves), shear waves, shock waves, surface waves, etc., or waves such as thermal waves (surface waves, (including guided waves, etc.), regardless of the presence or absence of transmission or reception focusing, transmission or reception steering, or transmission or reception apodization, the coordinate system for transmitting and receiving is different from the coordinate system for generating beamformed signals. Including cases, arbitrary beamforming can be performed with high precision and at high speed without performing interpolation approximation based on digital processing. In addition to improving the frame rate when displaying images of beamformed signals, high spatial resolution and high contrast can be obtained in terms of image quality, and furthermore, beamformed signals can be used to perform displacement, deformation, or Measurement accuracy can be improved by measuring temperature and the like. In the present invention, as described in methods (1) to (7), interpolation approximation processing may be performed in arbitrary beamforming wavenumber matching, and beamforming may be performed at a higher speed. In order to perform approximate wavenumber matching with high accuracy, it is necessary to appropriately oversample the received signal at the expense of an increase in the amount of calculation. In that case, it should be noted that the number of data for Fourier transform increases, unlike the case where signals at arbitrary positions can be selectively generated when interpolation approximation processing is not performed. Superposition processing and spectral frequency division of fast beamformed waves or beams, and fast beamforming in situations where received signals of unreceived beamforming are superimposed or spectrally frequency-divided enable various applications. do. The application of the present invention is not limited to this. High-speed processing is extremely effective in multidimensional imaging using multidimensional arrays. It should be noted that the Fourier transform or inverse Fourier transform processing performed in the above calculation algorithm preferably includes performing a dedicated fast Fourier transform or inverse fast Fourier transform. Moreover, the present invention is not limited to the above embodiments, and many modifications can be made within the technical concept of the present invention by those skilled in the art.

Objects to be measured include organic substances, inorganic substances, solids, liquids, gases, rheological objects, living creatures, celestial bodies, the earth, the environment, etc., and the range of applications is extremely wide. Non-destructive inspection (a recent topic is cheap and simple ultrasonic inspection of metals and plastics (especially Fiber-Reinforced Plastics using carbon fiber, glass fiber, etc.), etc.). , proposed ultrasonic observation of rubber, which is considered to have strong ultrasonic attenuation), diagnosis, resource exploration, generation and manufacturing of materials and structures, monitoring of various physical or chemical repairs and treatments. , contributed to the application of the clarified functions and physical properties, etc., in which the conditions that do not cause large disturbance to the measured object, are non-invasive, minimally invasive, and non-invasive were imposed. Sometimes precision is required. Ideally, the object may be observed in situ. For example, if 3D waves can be observed, 3D displacement vectors can be observed by Doppler. In the case of normal Doppler, it is necessary to orient the wave propagation direction in the displacement direction of the observation target, but the sensor should be directed to the observation target. It is possible to observe motion in any direction only by using the ultrasonic echo method (eg, ultrasonic echo method). Measurement techniques are also simplified. For example, when the eyelids are open, the ultrasound echo method is used when the eyelids are open or closed. We can reconstruct the viscoelasticity distribution of the eyeball and the intraocular pressure by observing the propagation and using the mathematical inverse problem. It is also possible to reconstruct blood viscosity, heart chamber pressure, and blood pressure by ultrasonic Doppler observation of blood flow vectors in heart chambers, blood vessels, fundus, and retina. In addition, not only light but also ultrasonic waves can be applied to various biometric authentication, including fingerprint identification (authentication) and iris authentication (including function observation such as dynamics and temperature, and various physical property observations). In addition, in some cases, observations can be made in situ while the observation target is in operation. For example, we can reconstruct the distribution of electrical physical properties such as electrical conductivity and dielectric constant of neural networks and electrical/electronic circuits that are running using mathematical inverse problems by observing the current density vector distribution based on the observation of the magnetic field distribution. , the temperature distribution can be observed to reconstruct the distribution of thermophysical properties, and the internal strain tensor and strain of the tire of a running car, the rubber tube through which a fluid flows, and the rubber of an electric wire through which an electric current flows. Ultrasound Doppler observation of the modulus tensor can be used to reconstruct the distribution of the viscoelastic modulus and internal pressure. Muscle motion vectors can be similarly observed to reconstruct viscoelasticity and tissue pressure. In addition, it is possible to monitor the generation/growth process without changing the physical conditions for observation while the material is being generated/grown. Paragraph 0094 describes an elastic wave observation method for observing the anisotropy of the elastic modulus (the observation can use ultrasound, MRI, OCT, etc.), but various other non-isotropic properties of physical properties are described. Isotropy can also be observed by observing waves related to each physical property (various electromagnetic waves, mechanical waves, heat waves, etc.) in the same way. In other words, by adjusting the position, number, directivity, etc. of the sources, we can observe the overlap of the waves to be observed (each electromagnetic wave, light, radiation, audible sound, ultrasonic wave, elastic wave, heat wave), can be observed and superimposed to control the direction of propagation. Alternatively, by adjusting the position, number, directivity, etc. of the sources, it is possible to generate and observe multiple purely independent waves to improve the accuracy of reconstructing physical properties (generate over-determined systems). ). For example, it can be applied to the design of devices (media, sources, etc.) that realize various surface waves and guided waves. Here, the object of observation is described as wave motion, but static or pseudo-static fields (strain tensor distribution, potential distribution, current density vector distribution, temperature distribution, etc.) may also be the object of observation. In some cases, static physical properties can be reconstructed by observing such fields.
In addition, when the SN ratio of waves to be directly sensed for observing those waves and fields is low, superposition and averaging of the waves are effective. For example, when observing the brain magnetic field of humans and animals (very weak and often buried in the surrounding magnetic field) with a SQUID meter and observing the current density vector, visual, auditory, and somatosensory Accuracy can be obtained by reconstructing the distribution of the current density vector by superimposing (adding and averaging) the induced brain magnetic field, and the distribution of electrical properties can be reconstructed. In some cases, the current density vector distribution is observed for each evoked and superimposed (averaged) to reconstruct the electrical property distribution. Also, in the case of a sudden onset such as epilepsy, the observed magnetic fields are superimposed and the observed time series are integrated to reconstruct them. In terahertz observation (electric field observation), averaging is performed to improve the SN ratio, and electrical properties may be reconstructed based on such superimposition. When signal components of a plurality of events are included in superimposition (additional averaging) or integration, their accumulation can be observed simultaneously, and random signal components are suppressed. It is possible to reconstruct them even in such superimposed signals. For example, the running structure of the neural network used during observation can be visualized by reconstructing the distribution of current density vectors and electrical properties. be done.
The frequency dispersion of physical properties can be reconfigured by changing the wave frequency or by realizing and observing wideband waves.
In addition, the action of the wave itself may be used to treat or repair the target, and beamforming may be performed on the response from the target at that time to observe the situation. In addition, satellite communication, radar, sonar, etc. implement beamforming to realize an informationally safe environment under energy saving, and accurate communication is also possible. The present invention is also effective in normal communications including ad-hoc communication devices and mobiles. It can also be applied to sensor networks. Real-time performance is required when the object is dynamic, but according to the present invention, it is possible to complete digital beamforming in a short period of time at high speed and with high accuracy.

<<Third Embodiment>>
Waves such as electromagnetic waves, light, mechanical vibrations, sound waves, and heat waves behave differently depending on their frequency, bandwidth, intensity, or mode. Many types of wave transducers have been developed so far, and imaging using transmitted waves, reflected waves, scattered waves, or the like of these waves has been performed. For example, it is well known that ultrasound, which has a high frequency among sound waves, is used in non-destructive inspection, medical treatment, and sonar. Radar also uses electromagnetic waves (microwaves, terahertz waves, infrared rays, visible light, radiation such as X-rays, etc.) with appropriate frequencies according to the observation target. The same is true for other waves.

In imaging using these waves, the distribution of amplitude data obtained through quadrature detection, envelope detection, or square-law detection is usually displayed as a gray image or color image in one, two, or three dimensions. . Also, in Doppler measurements using these waves, raw coherent signals are processed (ultrasound Doppler, radar Doppler, laser Doppler, etc.). Furthermore, in the field of image measurement, it is also well known that motion observation is performed using signals made incoherent through detection (cross-correlation processing, optical flow, etc.). In medical ultrasound and sonar, imaging using physically generated harmonics, chords, and difference tones is also performed.

Under such circumstances, the inventor of the present application is developing an ultrasound imaging technique for differential diagnosis of cancer lesions, sclerosis, and other lesions in human tissue. The inventor of the present application has been improving the resolution and efficiency of HIFU (High Intensity Focus Ultrasound) treatment along with improving the resolution of echo imaging and highly accurate measurement imaging of tissue displacement. We also perform their imaging based on the reception of echoes during ultrasound emission. These imagings are based on appropriate beamforming, and require appropriate detection methods and tissue displacement measurement methods.

For example, the inventors of the present application have used beamforming methods such as a lateral modulation method using cross beams, a spectral frequency division method, a method using many cross beams, and an over-determined system method, etc. , In particular, detection methods for multi-dimensional received signals include quadrature detection, envelope detection, and square-law detection. Displacement vector measurement methods include multi-dimensional autocorrelation, multi-dimensional Doppler, and multi-dimensional cross-spectrum phase gradient methods. , and devised a phase matching method, etc., and reported a technique for reconstruction imaging of (visco) shear elastic modulus distribution and thermophysical property distribution based on displacement and strain measurement (see Non-Patent Documents 13 and 30 reference). Some of them are already in clinical use. A recent report by the inventor of the present application is ITEC (International Tissue Elasticity Conference), IEEE Trans. On UFFC, Ultrasonic Study Group, and Acoustic Imaging Study Group.

In this regard, the inventors have focused on nonlinear imaging. In medical ultrasound, nonlinear imaging (so-called harmonic imaging) based on the result of physical action in the propagation process of ultrasound is currently performed. In the following, in particular, diagnostic and therapeutic applications of nonlinear ultrasound are discussed.

Harmonic imaging is based on the fact that wave components with high sound pressure intensities propagate at high velocities (usually explained as having a high bulk modulus for high intensity sound pressures). It is for imaging wave components. Contrast agents (ultrasound contrast agents) may be used in this harmonic imaging to enhance nonlinear effects in ultrasound propagation.

There is already a long history of clinical recognition of its effectiveness, such as the ability to perform blood flow imaging of capillaries (see Non-Patent Document 22). Doppler measurement using non-linear components (harmonic components) is also possible, and in the near future, we will report on multi-dimensional vector measurement realized by the present inventor in such a case. In Non-Patent Document 23, a so-called pulse inversion method is used to separate the fundamental wave.

In addition, tissue imaging has a history of being performed prior to blood flow imaging. Initially, harmonics were separated by filtering (see Non-Patent Document 24). Separated by the version method. If the transmit signal is broadband, the filtering method is limited because the fundamental and harmonic bands are covered. In addition, there is a report of separating a fundamental wave and harmonics represented by each dimension term based on the method of least squares in a multidimensional multiplicative polynomial consisting of power terms (see Non-Patent Document 25).

Recently, there are reports of applying harmonic components and chords to ultrasonic microscopes (see Non-Patent Document 26) and radiation pressure imaging (see Non-Patent Document 27). In addition, the nonlinear propagation and heat absorption are closely related, and HIFU is used by concentrating high-intensity ultrasonic waves on the focal position, including the case of causing cavitation (Non-Patent Document 28 etc.). Also, when the energy (or mode) is converted from ultrasonic waves to shear waves (for example, when sound waves are obliquely incident on the boundary where the acoustic impedance between soft tissue and bone changes greatly, or when shear occurs due to scattering) The high-frequency shear waves generated are susceptible to tissue absorption during propagation within the vicinity of the source (Girke).

Contrast agents used for the purpose of enhancing non-linear effects in HIFU treatment (see Non-Patent Document 29, etc.) are considered effective in these respects as well. 17 years ago, the inventor of the present application was the first in the world to refer to the effect of coagulating blood and occluding feeding arteries in the treatment of cancer lesions, and believes that this effect can also be obtained. . Recently, an applicator with a frequency band similar to that of clinical diagnostic probes has become available at low cost, but the inventor of the present application believes that it is necessary to develop a dedicated contrast agent. . The inventor of the present application considers at least fragile and non-fragile properties to be attractive properties, but in the near future it is possible to use the same properties or a mixture of several types for diagnostic purposes. be.

Waves are subject to attenuation while propagating, so the energy of a wave decreases as the propagation distance increases. Diffusing waves are also affected by diffusion. Under such circumstances, transmitted waves, reflected waves, or scattered waves reflect changes in impedance and the presence of reflectors or scatterers, and are used for their imaging and Doppler measurement. In their imaging, it is desirable that the signals contain high frequency components and be broadband to the extent possible or required, as in Doppler measurements.

Generally, however, high-frequency signal components are strongly affected by attenuation, and as the propagation distance increases, their energy is lost, the frequency of the signal is lowered, and the band becomes narrower. That is, imaging at locations far from the signal source suffers from lower signal-to-noise ratio (SNR) and lower spatial resolution. In Doppler measurement, its accuracy decreases. Reducing their effects due to attenuation is extremely important in an engineering sense.

Also, if a high-frequency signal that cannot be realized with a single signal source can be generated, imaging with higher resolution and Doppler measurement with higher precision will be possible. It may be possible to simply generate a high frequency signal. Generally, the influence of attenuation is strong in high-frequency components. For example, with a microscope that is susceptible to the influence of attenuation, it is desirable to be able to observe as deep as possible at a high frequency. Also, low-frequency imaging and measurement using low-frequency signals may be possible. Also, it may be possible to generate a low-frequency signal that cannot be realized with a single signal source. For example, the depth of interest can be deformed at low frequencies. In medical ultrasound imaging, nuclear magnetic resonance imaging, OCT, and laser applications, multiple signal sources are used to deform deep tissue at low frequencies (tissue elasticity).

For example, vibration is applied from multiple directions to generate vibration waves with a frequency lower than that frequency, or multiple ultrasonic beams are crossed at their focal positions to realize a low-frequency force source there. may generate low-frequency vibration waves, and the vibration waves generated may be ultrasonic waves (longitudinal waves), or the vibration waves generated may be shear waves (transverse waves). In some cases, it may be observed by ultrasound. It may be possible to adjust the direction of propagation of those generated waves. The ability to implement these signals theoretically or on a computational basis may also allow control over the waves generated. Furthermore, a detection method that can be easily implemented in a short period of time is also important in place of normal quadrature detection, envelope detection, and square-law detection.

Therefore, in view of the above points, the second object of the present invention is to enhance high frequency components that are generally relatively weak in intensity and lost high frequency components in arbitrary waves propagating from the measurement object. It is an object of the present invention to provide an imaging apparatus capable of improving spatial resolution and measurement accuracy by generating or newly generating images. In addition, the imaging apparatus enhances or simulates the nonlinear effect in the object to be measured, generates a new nonlinear effect when there is no nonlinear effect in the object to be measured, or virtually realizes and performs imaging. can be A third object of the present invention is to generate a high frequency signal that cannot be realized with a single wave source. A fourth object of the present invention is to realize a detection method that can be easily implemented in a short period of time.

In order to solve the above problems, an imaging apparatus according to one aspect of the present invention provides (i) nonlinear processing using a nonlinear device at an arbitrary position on a propagation path for arbitrary waves propagating from within a measurement target. (ii) receive by the transducer and generate an analog received signal, and perform analog non-linear processing on the analog received signal; and (iii) transducer a non-linear reception processing unit that performs at least one of the processes of receiving a digital received signal by generating an analog received signal by digitally sampling the analog received signal and performing digital non-linear processing on the digital received signal obtained by digitally sampling the analog received signal; An image signal generation unit that generates an image signal representing an image of a measurement target based on the reception signal obtained by the reception processing unit.

According to one aspect of the present invention, by performing nonlinear processing on a signal of a frequency in which the influence of attenuation is not a problem in arbitrary waves propagating from the object to be measured, generally relatively weak intensity Spatial resolution and measurement accuracy can be improved by enhancing or newly generating high-frequency components and lost high-frequency components. In addition, waves that arrive from arbitrary wave signal sources such as electromagnetic waves, light, mechanical vibrations, sound waves, or heat waves, and transmitted waves, reflected waves, or scattered waves of waves emitted from those signal sources Multiplication and exponentiation effects during wave propagation, and processing including analog and digital calculations of coherent signals obtained by detecting , by a transducer, can enhance nonlinear effects in the object to be measured. Alternatively, similar effects can be simulated, newly generated, or virtually realized.

For example, in imaging using coherent signals and Doppler measurements, compared to imaging using the original signal, we have achieved high-resolution imaging by using a wideband signal containing high-frequency components, and using the original signal Measurement of displacement, velocity, acceleration, strain, or distortion rate can be realized with higher precision than Doppler measurement. Incoherent signals are also subjected to similar processing for similar problems. As hardware, ordinary devices can be used. Of course, analog processing (circuitry) is faster than digital processing (circuitry). Calculators and devices with arithmetic functions (FPGA, DSP, etc.) can be used when performing calculations including higher-order calculations and in terms of a wide degree of freedom.

In particular, it is robust against the effects of attenuation in the wave propagation process, and can generate high-frequency components that cannot be realized with a single signal source, enabling high-resolution imaging and high-precision Doppler measurement. It is also possible to generate high-frequency signals that cannot be physically realized with a single signal source. If a plurality of 100 MHz ultrasonic transducers are used, it is possible to physically generate ultrasonic waves that are twice as high as the number of ultrasonic transducers, thereby achieving a high frequency that cannot be generated by ordinary transducers. Alternatively, it may be possible to simply generate a high-frequency signal. According to the present invention, such high frequencies can also be realized by computation. Therefore, it is possible to generate high-frequency waves and signals that cannot be physically realized. Similarly, low-frequency imaging and measurements using low-frequency signals can be realized. It is also possible to generate low frequency signals that are physically impossible with a single signal source. These signals can also be implemented theoretically or on a computational basis to control the waves generated.

For example, in an ultrasonic microscope, a high-frequency ultrasonic wave (signal) of several hundred MHz is applied to generate an ultrasonic wave with a frequency higher than that determined by the sound source, and the ultrasonic wave is robust against attenuation. Therefore, it is possible to realize an ultrasonic microscope capable of imaging with higher resolution and Doppler measurement with higher accuracy than usual. Also, low-frequency imaging and measurements using low-frequency signals are possible. In addition, when measuring the deformability of tissue, for example, the deep part of the object can be deformed at a low frequency. In medical ultrasound imaging, nuclear magnetic resonance imaging, OCT, laser applications, etc., multiple signal sources are used to deform deep tissue at low frequencies. The same applies to other imaging devices and Doppler devices. In addition, effects such as warming, heating, cooling, freezing, welding, heat treatment, cleaning, or repair are improved, and high resolution is also possible at that time. Similar effects can be obtained for incoherent signals after various detections.

In terms of signal processing technology, quadrature detection and envelope detection can be easily performed. For example, if the present invention is applied to a deflected beam or a deflected wave, IQ signals obtained by quadrature detection with respect to all coordinate axes can be obtained, facilitating envelope detection. Further, when the present invention is applied to cross beams, orthogonally detected IQ signals are obtained on each coordinate axis. Alternatively, it becomes possible to measure the strain factor tensor. Of course, in imaging, square-law detection can also be performed as detection processing.

Waves arriving from signal sources of arbitrary waves such as electromagnetic waves, light, mechanical vibrations, sound waves, or heat waves, and transmitted waves, reflected waves, or scattered waves emitted from those signal sources are transducers. Imaging and Doppler measurement using coherent signals obtained by detection are widely performed using suitable frequencies for each medium in radar, sonar, non-destructive inspection, diagnosis, and the like. Wave motions emitted from a signal source are also applied to heating, heating, cooling, freezing, welding, heat treatment, cleaning, or repair. Furthermore, in recent years, image measurement such as movement using incoherent signals has come to be performed, and various imaging and measurement are performed based on image processing and signal processing. The present invention is effective in all of these, and the applicability and marketability of the present invention are very high.

FIG. 37 is a block diagram showing an example configuration of an imaging apparatus according to the third embodiment of the present invention. This imaging device captures an image of an object to be measured based on arbitrary waves such as electromagnetic waves, light, mechanical vibrations, sound waves, or heat waves coming from the object to be measured, or measures physical quantities such as displacement in the object to be measured. It is a non-destructive measurement device.

As shown in FIG. 37, the imaging device includes at least one transducer 110 that is both transmitting means and receiving means, and an imaging device body 120 . Transducer 110 may be capable of generating and receiving arbitrary waves such as electromagnetic waves, light, mechanical vibrations, sound waves, or heat waves. In that case, the transducer 110 can transmit arbitrary waves to the measurement target 1 and receive reflected waves reflected within the measurement target 1 or scattered waves scattered within the measurement target 1 . For example, when the arbitrary wave is an ultrasonic wave, it is possible to use an ultrasonic transducer that transmits the ultrasonic wave according to the drive signal and receives the ultrasonic wave to generate a reception signal. It is well known that there are different ultrasonic elements (PZT, piezoelectric polymer, etc.) and different transducer structures depending on the application.

Historically, narrow-band ultrasonic waves have been used in blood flow measurement, but the inventors of the present application have found that displacement and strain (including static cases) of soft tissues and shear waves that have been put into practical use in recent years We have been the first in the world to realize the use of broadband transducers for (echo) imaging, including for propagation (velocity) measurements. The same is true for HIFU treatment, and continuous waves are sometimes used, but the inventors of the present application are developing devices using broadband devices in order to achieve high-resolution treatment. When using high-intensity ultrasonic waves, the tissue is stimulated in a range that does not cause a heating effect, and a force source may be generated in the measurement object 1 as described above, and a transducer for (echo) imaging is used. sometimes Heat therapy, force generation and (echo) imaging may be performed simultaneously. The same is true for other wave sources and transducers. There are two types of transducers: contact type and non-contact type, and each wave is properly impedance-matched before use.

Alternatively, the transducer 110 may be a transmitting transducer that generates arbitrary waves and a receiving transducer (sensor) that receives arbitrary waves. In that case, the transmitting transducer transmits an arbitrary wave to the measurement object 1, and the sensor transmits a reflected wave or a scattered wave reflected in the measurement object 1, or a transmitted wave transmitted through the measurement object 1. Waves, etc. can be received.

For example, if the arbitrary wave is a heat wave, a heat source that is not intentionally generated such as sunlight, lighting, or metabolism in the body may be used, but an infrared warmer, heater, etc. Ultrasonic transducers that transmit ultrasonic waves for heating that are relatively stationary and often controlled according to drive signals (sometimes generate a force source within the measurement object 1), electromagnetic transducers, lasers, etc. is also used. In addition, infrared sensors that receive heat waves and generate reception signals, pyroelectric sensors, microwave and terahertz wave detectors, temperature sensors such as optical fibers, ultrasonic transducers (temperature-dependent ultrasonic sound velocity, volume change, etc.) A nuclear magnetic resonance signal detector (detects temperature using chemical shift of nuclear magnetic resonance) can be used. For each wave, a suitable receiving transducer is used.

The transducer 110 may actively generate waves including harmonics when actively generating waves according to the drive signal. For example, the transducer 110 produces waves according to the non-linear characteristics of the wave source or the circuitry of the transmitter 121 that drives it. Also, transducer 110 may have a single transmit or receive surface, or multiple transmit or receive surfaces. A transmission surface of the transducer 110 may be provided with a nonlinear device 111 that performs nonlinear processing on the generated arbitrary wave. A receiving surface of the transducer 110 may be provided with a nonlinear device 111 that performs nonlinear processing on arbitrary waves propagating from within the object 1 to be measured. The nonlinear device 111 does not necessarily have to be in contact with the transmitting surface or receiving surface of the transducer 110, and may be provided at any position along the arbitrary wave propagation path.

Also, between the object 1 to be measured and the transmitting or receiving surface of the transducer 110, a working device 112 such as a filter (such as a spectrometer), a shield, an amplifier or an attenuator may be provided. When using the nonlinear device 111 , the working device 112 may be provided on both front and rear sides of the nonlinear device 111 . Transducer 110, nonlinear device 111, and working device 112 may be separate or integral in combination.

FIG. 37 also shows the case where the wave source is provided within the object 1 to be measured, or it may be directly controllable by the controller 133 . In addition, the wave motion generated by the transducer 110 may be concentrated using a lens or the like, or a plurality of transducers 110 may be used to perform focus transmission or the like to generate a wave motion source (mechanical wave or Sources of heat waves, or using mechanical waves or electromagnetic waves to generate new electromagnetic waves, such as magnetic materials that may be contrast agents, or to the physical effects and properties between waves (including the case of controlling the strength and propagation direction of waves by stimulating the waves).

Of course, the measurement object 1 may originally have a wave source (for example, the electrical activity of the brain and heart is a current source, and the heart is a force source). In addition, there are cases where the wave source is controllable, there are cases where the wave source cannot be controlled, there are cases where the measurement target 1 is observed in situ, or those wave sources themselves are the imaging target or the measurement target. Sometimes it is. Alternatively, such a wave source may originally exist outside the measurement target 1, and may be treated in the same way and become a measurement target. A nonlinear device 111 or an action device 112 may be appropriately provided between such a wave source and the object 1 to be measured.

Furthermore, in at least a part of the measurement object 1, a contrast medium such as microbubbles (nonlinear enhancement Agent) 1a may be injected. As the contrast medium 1a, a medium having an affinity for a target lesion, fluid, or the like in the measurement object 1 may be used. Thus, a transducer receiving waves may receive waves generated by multiple wave sources.

The transducer 110 is supplied with a driving signal from the imaging apparatus main body 120 and/or outputs a reception signal to the imaging apparatus main body 120 by wire or wirelessly. In the case of wireless, a wireless receiver and/or wireless transmitter is provided within the transducer 110 and a wireless transmitter and wireless receiver are also provided within the imaging apparatus body 120 .

In Part A, the imaging apparatus main body 120 includes a transmitter 121, a receiver 122, a filter/gain adjustment section 123, a nonlinear element 124, a filter/gain adjustment section 125, a detector 126, and an AD (Analog- to-digital converter 127 and storage device 128 . In Part B, the imaging apparatus main body 120 includes a reception beamformer 129, a calculation unit 130, an image signal generation unit 131, a measurement unit 132, a control unit 133, a display device 134, and an analog display device 135. may contain The control unit 133 controls each unit of the imaging apparatus body 120 .

If multiple transducers 110 are used, there may be as many Part A channels as there are transducers 110 . A case in which these transducers 110 form an array will also be described later. As shown in FIG. 37, when a multi-channel part A is provided, the received signal output from the multi-channel part A storage device 128 may be supplied to the part B receive beamformer 129. . Alternatively, each part B may be connected in cascade to each part A and processed independently, in which case the received signal output from each channel's part A storage device 128 is stored in each channel. is supplied to the receive beamformer 129 of Part B of the . It should be noted that multiple transducers 110 may be associated with different types of waves, in which case nonlinear effects of different types of waves may be observed simultaneously, and nonlinear effects in the same wave may be observed at the same time. We sometimes observe nonlinear effects between different kinds of waves.

In Part A, transmitter 121 to detector 126 may be configured by analog circuits, or at least some of them may be configured by digital circuits. In Part B, the receiving beamformer 129 to the control unit 133 may be configured by digital circuits, or a central processing unit (CPU) and software for causing the CPU to perform various processes are recorded. It may be configured by a recording medium. A hard disk, flexible disk, MO, MT, CD-ROM, DVD-ROM, or the like can be used as the recording medium. At least part of the reception beamformer 129 to the control unit 133 may be composed of analog circuits.

The transmitter 121 includes a signal generator such as a pulser that generates a drive signal according to a trigger signal supplied from the controller 133 . The control unit 133 may control the frequency, carrier frequency, bandwidth, transmission signal strength (apodization), or the waveform or shape of a pulse wave, burst wave, or the like. The control unit 133 may set the timing or delay time of the trigger signal for each channel. Alternatively, the timing of the trigger signal output from the control unit 133 is kept constant for all channels, and the transmitter 121 may further include a delay element that delays the trigger signal according to the delay time set by the control unit 133. .

The transmitter 121 applies the generated drive signal to the transducer 110 to cause the transducer 110 to generate arbitrary waves. For example, the transmitter 121 includes a drive signal amplifier (which may also serve as apodization) in order to adjust the intensity of transmitted waves and the intensity of generated harmonics, and a delay time set by the control unit 133 . Further elements may be included. A drive signal containing harmonics may also be generated and used. Various waves are generated and used, such as generating a chirp wave when apodization is performed instead of resonance or when forced vibration is applied. When the driving signals generated by the multi-channel transmitter 121 are applied to the plurality of transducers 110, the setting of the delay time by the control unit 133 enables focusing and steering of the transmission beam and transmission of plane waves. (A plane wave is a narrow-band wave in the direction perpendicular to the direction of propagation, and is effectively broadband).

In addition, the transmitter 121 may further include nonlinear elements (analog devices such as transistors, diodes, and nonlinear circuits, or digital devices such as nonlinear arithmetic units) in which nonlinear effects are similarly set. Pre-prepared frequencies, carrier frequencies, bandwidths, apodizations, delays, and non-linear effects may be set, but may be controlled by the operator via the control unit 133, and They may be adaptively determined and controlled in accordance with the observation situation in the computing unit 130 .

When driving a plurality of transducers 110, the transmitter frequency, carrier frequency, bandwidth, apodization, delay element, and nonlinear element of each channel are controlled, and these patterns are set in advance. In some cases, the pattern may be controlled by the operator via the control unit 133, and the pattern may be adaptively determined and set in the calculation unit 130 according to the observation situation. be.

The receiver 122 may include, for example, an amplifier that amplifies the received signal or an attenuator that attenuates the received signal (which can also serve as an apodization or filter), and may further include a delay element whose delay time is set by the controller 133 . The receiver 122 may further include nonlinear elements (analog devices such as transistors, diodes, and nonlinear circuits, or digital devices such as nonlinear computing units) that produce nonlinear effects. They may be configured similar to those of transmitter 121, including when waves are received by multiple transducers 110. FIG. The receiver 122 amplifies the received signal generated by the transducer 110 that has received the arbitrary wave and outputs the amplified received signal to the filter/gain adjuster 123 and AD converter 127 .

The filter/gain adjuster 123 includes a filter that limits the frequency band of the received signal, or an amplifier or attenuator that adjusts the gain of the received signal. Filter/gain adjustment section 123 adjusts the frequency band or gain of the received signal and outputs the received signal to nonlinear element 124 .

The nonlinear element 124 includes, for example, analog devices such as transistors, diodes, and nonlinear circuits, and applies analog nonlinear processing to the received signal. This nonlinear processing may be a power operation for at least one frequency component signal included in the received signal, or may be a multiplication operation for a plurality of frequency component signals included in the received signal ( Hall effect elements, etc. can be used).

The filter/gain adjuster 125 includes a filter that limits the frequency band of the received signal, or an amplifier or attenuator that adjusts the gain of the received signal. Filter/gain adjustment section 125 adjusts the frequency band or gain of the received signal and outputs the received signal to detector 126 and AD converter 127 .

The filter/gain adjustment units 123 and 125 and the nonlinear element 124 may be set to those prepared in advance by the control unit 133, but they are controlled by the operator via the control unit 133. In addition, they may be adaptively determined and controlled in the calculation unit 130 according to the observation situation. When driving a plurality of transducers 110, they are independently controlled in each channel. may be controlled, and these patterns may be adaptively determined and set in the calculation unit 130 according to the observation situation.

For example, the detector 126 generates an analog image signal by performing envelope detection processing, square-law detection processing, or the like on the received signal when reception beamforming is not performed. Also, displacement measurement may be performed through quadrature detection. The analog display device 135 displays an image of the measurement target 1 or the wave source based on the image signal generated by the detector 126 and the measurement result.

AD converter 127 selects the received signal output from filter/gain adjustment section 125 when analog nonlinear processing is applied to the received signal, and receiver 122 selects the received signal output from the The AD converter 127 digitally samples the analog received signal to convert it into a digital received signal. A digital received signal obtained by the AD converter 127 is output to the storage device 128 . The storage device 128 is configured by a memory such as RAM, for example, and stores the received signal.

The received signals stored in storage device 128 are provided to receive beamformer 129 . It should be noted that while the signals are being processed by the receive beamformer 129, the signals being processed may be temporarily stored in the storage device 128 or the external storage device 140 and read out as needed. When a single or a plurality of transducers 110 are used, the receive beamformer 129 may separate harmonics by a pulse inversion method, a polynomial fitting method, or the like (the same processing is performed in the calculation unit 130). sometimes called).

Also, when multiple transducers 110 are used, the receive beamformer 129 performs receive beamforming processing on the received signals supplied from the multi-channel storage device 128 . For example, the reception beamformer 129 delays the reception signals of a plurality of channels according to the delay time set by the control unit 133 and performs phasing processing, and then performs addition or multiplication operations to synthesize the reception signals and focus them. generates a new received signal in which is narrowed down.

Alternatively, if the receiver 122 includes a delay element, the multiple channel receiver 122 may delay each received signal according to the delay time set by the controller 133 . The receive beamformer 129 synthesizes received signals by adding or multiplying the received signals of multiple channels. During receive beamforming, the receive beamformer 129 may perform apodization.

In addition, by mounting a (multi-dimensional) fast Fourier transformer on the imaging device main body 120 and obtaining the spectrum of the received signal, filtering and beam or wave characteristics (frequency, carrier frequency, bandwidth , frequency or carrier frequency in any direction, bandwidth in any direction, waveform, beam shape, steering direction or direction of propagation, etc.). By spectral frequency division (see Non-Patent Document 30), it is possible to obtain multiple received signals (corresponding to pseudo multiple different beamforming) from a single received signal, and to perform nonlinear processing on these. be. These processes are sometimes performed on signals that have undergone nonlinear processing.

In this imaging apparatus, the plurality of wave signals (with or without nonlinear effects) generated using a single or multiple transducers are stored in the storage device 128 or the external storage device 140, The reception beamformer 129 or the calculation unit 130 reads the results and performs addition (superposition, linear processing) and multiplication (nonlinear processing) calculations, which may be used for imaging and various measurements. In that case, phasing processing is performed appropriately.

The arithmetic unit 130 mainly performs digital non-linear processing on the digital reception signal output from the reception beamformer 129 . This nonlinear processing may be an exponentiation operation on at least one frequency component signal included in the received signal, or may be a multiplication operation on a plurality of frequency component signals included in the received signal. As described above, the calculation unit 130 may also serve as the reception beamformer 129 . In any event, while the signals are being processed, the signals being processed may be temporarily stored in storage device 128 or external storage device 140 and read out as needed.

Here, the transducer 110 to the action device 112 and the receiver 122 to the calculation unit 130 perform nonlinear processing on the received signal obtained by receiving the arbitrary wave propagating from the measurement object 1 or the arbitrary wave. It constitutes a non-linear reception processing unit that applies In the nonlinear reception processing unit, at least one of the nonlinear device 111, the nonlinear element 124, and the arithmetic unit 130 receives an arbitrary wave propagating from the measurement object 1 or a received signal obtained by receiving the arbitrary wave. Non-linear processing is applied to the In addition, nonlinear effects may be obtained as described above.

That is, the nonlinear reception processing unit (i) performs nonlinear processing using the nonlinear device 111 at an arbitrary position on the propagation path with respect to the arbitrary wave propagating from the object 1 to be measured, and then receives the wave by the transducer 110. A process of generating a received signal may be performed. In addition, the nonlinear reception processing unit (ii) receives the arbitrary wave propagating from the object 1 to be measured by the transducer 110 to generate an analog reception signal, and converts the analog reception signal into, for example, an analog nonlinear An analog non-linear processing may be performed using the element 124 . In addition, the nonlinear reception processing unit (iii) receives the arbitrary wave propagating from the object 1 to be measured by the transducer 110, generates an analog reception signal, and digitally samples the analog reception signal. A digital received signal may be subjected to digital non-linear processing using, for example, the digital computing unit 130 . In addition, nonlinear effects may be obtained as described above.

The image signal generating unit 131 and the measuring unit 132 select the received signal output from the arithmetic unit 130 when performing digital nonlinear processing on the received signal, and select the received signal when not performing digital nonlinear processing on the received signal. A received signal output from the beamformer 129 is selected.

The image signal generator 131 generates an image signal representing the image of the measurement target 1 based on the received signal obtained by the nonlinear reception processor. Alternatively, the image signal generator 131 may generate an image signal based on a received signal obtained by nonlinear processing as well as a received signal that has not undergone nonlinear processing. Further, the image signal generator 131 may generate an image signal representing the image of the measurement target 1 by selectively using the received signal obtained when nonlinear processing is not performed. For example, the image signal generator 131 generates an image signal by performing envelope detection processing, square-law detection processing, or the like on the received signal. The display device 134 generates an image signal representing the image of the measurement target 1 based on the image signal generated by the image signal generating section 131 .

The measurement unit 132 measures displacement or the like in the measurement target 1 using at least one of the plurality of signals obtained by nonlinear processing. For example, in observing the propagation of a mechanical or electromagnetic wave, the measurement unit 132 measures the particle displacement or particle velocity caused by arbitrary wave propagation of its own wave or another wave based on the measured displacement. In that case, the image signal generation unit 131 generates an image signal representing wave propagation based on the particle displacement or particle velocity measured by the measurement unit 132 . When a plurality of waves arrive, the waves are separated in advance, or the measurement is performed through analog processing or digital processing after reception.

Alternatively, in observing the propagation of thermodynamic waves, the measurement unit 132 may use, as the transducer 110, an infrared sensor, a pyroelectric sensor, a microwave or terahertz wave detector, a temperature sensor such as an optical fiber, an ultrasonic transducer (ultrasonic transducer). Temperature changes are detected using the temperature dependence of sound velocity and volume changes, etc.), or heat waves are measured using a nuclear magnetic resonance signal detector (temperature is detected using the chemical shift of nuclear magnetic resonance). You can In that case, the image signal generation unit 131 generates an image signal representing thermodynamic wave propagation based on the heat waves measured by the measurement unit 132 . The image signal generated by the image signal generating section 131 and the measurement data obtained by the measuring section 132 can be stored in the external storage device 140 .

In the above, the nonlinear reception processing unit obtains the result of the exponentiation operation by the nonlinear processing for the arbitrary wave propagating from the object 1 to be measured, or obtains the result of the exponentiation operation based on the arbitrary wave by the nonlinear processing being the exponentiation operation. As a result of chords, difference tones, and overtones, a received signal having a higher frequency or a lower frequency than the received signal obtained without non-linear processing may be obtained. Also, the non-linear processing may be a multiplication operation. The non-linear processing may be a high-order non-linear processing, and due to its effect, the results of power operations and multiplication operations are often obtained.

As a result, when the arbitrary wave has a plurality of different frequency components, the received signal has a wider band than the received signal obtained without nonlinear processing. Alternatively, the high-frequency received signal becomes a harmonic signal with higher frequency, higher spatial resolution, lower sidelobes, or higher contrast than the received signal obtained when nonlinear processing is not performed. . Alternatively, the frequency-lowered received signal becomes a signal in a band including direct current obtained by substantially quadrature detection of the harmonic signal. The image signal generator 131 generates an image signal based on at least one signal obtained by nonlinear processing.

Alternatively, a plurality of arbitrary waves propagating from within the object 1 to be measured are propagated in the direction of propagation, steering angle, frequency, carrier frequency, pulse shape, beam shape, frequency in any direction, carrier frequency, Alternatively, when at least one of the bandwidths is different, the nonlinear reception processing unit performs at least one of the above processes (i) to (iii) on a plurality of arbitrary waves that arrive overlappingly. Also good. The image signal generator 131 generates an image signal based on the reception signal obtained by the nonlinear reception processor.

Here, before receiving the plurality of arbitrary waves, the nonlinear reception processing unit passes the plurality of arbitrary waves through at least one of an analog delay device and an analog storage device as the working device 112, thereby obtaining a plurality of arbitrary waves. Wave motions may overlap at each position within the object 1 to be measured. This is so-called aberration correction.

In addition, the nonlinear reception processing unit obtains the result of the exponentiation operation by the nonlinear processing of the superimposition of a plurality of arbitrary waves propagating from within the measurement object 1, or the nonlinear processing is the exponentiation operation, Based on a plurality of arbitrary waves, as a result of chords, difference tones, and overtones, a received signal having a higher frequency or a lower frequency than the received signal obtained without nonlinear processing may be obtained. Also by this, the same effect as the above can be obtained. The image signal generator 131 generates an image signal based on at least one signal obtained by the nonlinear reception processor.

Alternatively, a plurality of arbitrary waves propagating from within the measurement target 1 are transmitted in the measurement target 1 in the propagation direction, steering angle, frequency, carrier frequency, pulse shape, beam shape, and frequency or carrier in any direction. When at least one of the frequencies or bandwidths is different, the nonlinear reception processing unit performs at least one of the above processes (i) to (iii) for a plurality of arbitrary waves that arrive at the same time. At any time after receiving multiple random waves, the received signal may be separated into multiple signals using analog or digital devices or based on analog or digital signal processing. The image signal generator 131 generates an image signal representing the image of the measurement target based on at least one of the plurality of signals separated by the nonlinear reception processor. Non-linear operations are performed to obtain multiplicative effects. Also, analog or digital aberration correction may be performed and then the signals added again to obtain a power effect.

Alternatively, a plurality of arbitrary waves propagating from within the measurement target 1 are transmitted in the measurement target 1 in the propagation direction, steering angle, frequency, carrier frequency, pulse shape, beam shape, and frequency or carrier in any direction. Waves that arrive non-overlapping, waves that are shielded from non-overlapping using working devices 112, and devices (analog or digital) or The nonlinear reception processing unit may perform at least one of the above processes (i) to (iii) on at least one of the waves separated using analog or digital signal processing. . The image signal generator 131 generates an image signal based on the reception signal obtained by the nonlinear reception processor.

Here, before receiving the plurality of arbitrary waves, the nonlinear reception processing unit passes the plurality of arbitrary waves through at least one of an analog delay device and an analog storage device as the working device 112, thereby obtaining a plurality of arbitrary waves. Wave motions may be superimposed at each position within the object 1 to be measured. This is so-called aberration correction.

Alternatively, the non-linear reception processing unit passes the analog received signal through at least one of an analog delay device and an analog storage device, delays the digital received signal by digital calculation, or delays the digital received signal is passed through a digital storage device, a plurality of arbitrary waves may be superimposed at each position within the object 1 to be measured.

In addition, the nonlinear reception processing unit obtains the result of the exponentiation operation by nonlinear processing for each of the plurality of arbitrary waves propagating from the object 1 to be measured, or the nonlinear processing is exponentiation operation, whereby the plural As a result of chords, difference tones, and overtones based on each of the arbitrary waves, a received signal having a higher frequency or a lower frequency than the received signal obtained without nonlinear processing may be obtained. . Also by this, the same effect as the above can be obtained. The image signal generator 131 generates an image signal based on at least one signal obtained by the nonlinear reception processor.

Alternatively, the nonlinear reception processing unit obtains the result of the multiplication operation by performing nonlinear processing on each of the plurality of arbitrary waves propagating from inside the measurement object 1, or the nonlinear processing is a multiplication operation, whereby the plurality of arbitrary waves are obtained. Based on the waves, chords and differences or overtones may result in a received signal that is higher or lower in frequency than the received signal obtained without non-linear processing.

As a result, when a plurality of arbitrary waves have a plurality of different frequency components, the received signal has a wider band than the received signal obtained without nonlinear processing. Alternatively, the high-frequency or low-frequency received signal has higher spatial resolution, lower sidelobes, or higher contrast than the received signal obtained when nonlinear processing is not performed, and at least any one Approximately quadrature detection is performed in one direction, resulting in a signal in a band containing direct current and harmonic frequencies in at least one other direction. The image signal generator 131 generates an image signal based on at least one signal obtained by nonlinear processing.

In the above, the image signal generation unit 131 performs arbitrary detection processing on at least one of the plurality of signals obtained by nonlinear processing, or performs arbitrary detection processing on a superimposition of a plurality of signals, Alternatively, an image signal may be generated by superimposing a plurality of signals after performing arbitrary detection processing on the plurality of signals.

<<Fourth Embodiment>>
Next, a fourth embodiment of the invention will be described. FIG. 38 is a block diagram showing a configuration example of an imaging apparatus according to the fourth embodiment of the present invention and modifications thereof. The imaging apparatus according to the fourth embodiment and its modification is an apparatus that drives a plurality of transducers 110 or a transducer array to generate waves or receives waves by them to perform imaging (see FIG. 38 indicates a transducer array), and the components can be those having the same performance as the components in the third embodiment.

In the imaging apparatus according to the fourth embodiment shown in FIG. 38(a), as in the case of using a plurality of transducers 110 or transducer arrays in the imaging apparatus according to the third embodiment shown in FIG. A transducer 110 or element is connected to a plurality of transmitters 121 or receivers 122, respectively. However, in the imaging apparatus main body 120a, a plurality of transmitters 121 or receivers 122 are provided in one part A'.

A summing unit sums (linearly ), or the multiplication unit performs analog processing of multiplication (non-linear processing, Hall effect elements, etc. are used). As a result, receive beamforming is performed, so in Part B, the receive beamformer 129 (FIG. 37) is not required.

A digital received signal obtained through the AD converter 127 is then stored in the storage device 128 . The part B of the imaging apparatus main body 120a is controlled by the control unit 133 to obtain all the nonlinear effects that can be achieved by the third embodiment based on the received signal, thereby achieving the same as the third embodiment. Similarly, imaging and measurement imaging are performed. Note that wiring from the control unit 133 to the receiver 122 and the like is omitted in FIG.

In the fourth embodiment, similarly to the transmitter and receiver in the third embodiment, it is possible to give a delay to the driving signal or the receiving signal of each transducer, and to perform focusing and steering for transmission or reception. Processing is also possible. In the fourth embodiment, one AD converter 127 and one storage device 128 are required compared to the third embodiment which requires the same number of AD converters 127 and storage devices 128 as the number of channels. Since it is only necessary to provide them, the device can be simplified.

On the other hand, in the imaging apparatus according to the modification of the fourth embodiment shown in FIG. It is provided outside the vessel 122 . Unlike the imaging device shown in FIG. 38(a), the imaging device shown in FIG. The signal is received by the receiver 122 after the processing unit performs addition (linear processing) analog processing, or the multiplication processing unit performs multiplication (nonlinear processing) analog processing. Therefore, since it is sufficient to provide one transmitter 121 and one receiver 122, the device can be greatly simplified, and the same nonlinear effect as in the third embodiment can be obtained.

The imaging apparatus according to the third embodiment shown in FIG. 37, the imaging apparatus according to the fourth embodiment shown in FIG. 38(a), and the imaging apparatus according to the modification of the fourth embodiment shown in FIG. 38(b) Devices or other types of imaging devices or components thereof may also be used simultaneously. For example, coherent or incoherent image signals and measurements obtained using multiple types of equipment can be individually displayed, displayed side by side at the same time, superimposed or multiplied. You can also display things. Basically, those obtained from received signals at the same time or the same time phase can be targeted. Even in the same imaging apparatus, the same processing may be performed when a plurality of image signals or measurement results are obtained using signals received at the same time or at the same time phase. The target signal is an analog signal or a digital signal after phase adjustment, and the addition or multiplication is performed by analog processing (using a Hall effect element, etc.) or digital processing (using a calculator, arithmetic unit, etc.). .

The imaging apparatus according to the first or fourth embodiment of the present invention uses analog calculators, digital calculators, calculators of various devices, or similar devices (FPGA, DSP, etc.) to perform nonlinear calculation on signals. on the basis of As will be described in detail later, non-linear operations are mainly to obtain the effects of exponentiation and multiplication, but the operations themselves are not limited to this, and may be higher-order calculations having other non-linear characteristics. These effects can be extracted or isolated through polynomial fitting, spectral analysis, pulse inversion techniques, numerical calculations, or signal processing. Non-linear calculations may be applied not only to signals but also to waves, and before reception, waves are extracted using devices (time or space, or their frequency filters, spectroscopes, etc.) or can be separated. A dedicated device may be available.

As described above, this imaging apparatus includes the arbitrary wave transducer 110, the transmitter 121, and the receiver 122, and also includes the nonlinear device 111, the nonlinear element 124, or the arithmetic unit 130. Data storage devices (memory, hard disk, photo, CD-RW, or other recording media), display devices, etc. are provided according to the requirements. The nonlinear processing of the present invention can be applied to an existing device that does not have the nonlinear device 111, the nonlinear element 124, the arithmetic unit 130, or other nonlinear devices. Non-linear processing can also be implemented in addition to devices that do.

Waves transmitted from the transmitter 121 or the transducer 110, which serves as a wave source, may be pulse waves, burst waves, or coded waves such as phase-modulated waves, enabling imaging and measurement with spatial resolution. be. However, when measurement is possible without requiring spatial resolution, this is not the case and continuous waves may be used.

The waves generated are determined by the conversion characteristics of the electrical signal (drive signal) to waves in the transducer 110, and appropriately designed devices and drive signals are used. For example, in the case of light waves, various light sources (coherent or incoherent, light emitting diodes (LEDs), light-mixing LEDs, (tunable wavelength) lasers, or optical oscillators, etc.) are used, and in the case of sound waves, an electrical source is used as the source. Acoustic transducers, vibrators, etc. are used. In the case of vibration waves, an actuator-based vibration source is used, and in the case of thermal waves, a heat source or the like is used. Thus, in this embodiment, transducers 110 that generate various waves can be used.

The transducers 110 to be used include representative transducers for the various waves described above, and it is also possible to positively use transducers that are not normally used due to their nonlinear characteristics. Normally, when a high pressure is applied to an ultrasonic element, ultrasonic waves containing harmonics are generated due to nonlinear phenomena. Harmonic imaging, which extracts the nonlinear components generated in the medium, is performed by the pulse inversion method. In other cases, the harmonic content is filtered out and only the fundamental band signal is used.

Also in this embodiment, nonlinear waves may be positively generated and used. That is, when a nonlinear characteristic appears during transmission, the transmission wave includes harmonics, and this may be effectively applied in the present invention. On the other hand, a wave that does not contain nonlinear components is also generated to search for nonlinear phenomena in the object to be measured.

Moreover, in waves having nonlinear components, nonlinear phenomena may also occur in harmonics. If the transmitted wave originally contains harmonics or if there are multiple intersecting waves (frequency, carrier frequency, pulse shape, beam shape, or frequency seen in each direction, carrier frequency, or bandwidth (i.e., wave parameters other than propagation direction and steering angle are different), as described below, include pulse inversion, spatio-temporal filtering, spectral filtering, or polynomial fitting for the received signal. The present invention may be implemented after separation is achieved by analog processing or digital processing such as signal processing, or the present invention may be implemented without separation. In addition, in the propagation process, shielding objects such as obstacles, filter devices, spectroscopes (time or space, or those of their frequencies), physical stimuli are applied to change the refractive index of the medium (optical In some cases, the waves are separated in advance by using a switch, etc., and each wave is received. If the source of each wave can be controlled, each wave can be generated independently and each can be observed.

In addition, after the wave is generated in the transducer 110 and before the wave propagates to the object to be measured, by using a device that directly causes a nonlinear phenomenon in the wave, the wave containing the nonlinear component can be transferred to the object to be measured. may be propagated to It is also possible to use devices that induce non-linear phenomena after or during propagation in the measurement object. Waves or signals may also be combined, mixed, etc. to obtain a multiplicative effect.

For example, with regard to light, (i) nonlinear optical elements (for example, optical harmonic generation devices used for wavelength conversion of laser light to a short wavelength region, etc.), (ii) optical mixing devices, and (iii) optical parametric generation , stimulated Raman scattering, coherent Raman scattering, stimulated Brillouin scattering, stimulated Compton scattering, or devices that produce optical parametric effects such as four-wave mixing; (iv) multiphoton transitions such as ordinary Raman scattering (spontaneous emission Raman scattering) (v) a device that produces a nonlinear refractive index change; and (vi) a device that produces an electric field dependent refractive index change. Couplers, optical fibers, etc. can also be used effectively. Multi-point observation (multi-channel) is also possible, and it is also suitable for signal processing.

When using light, it relates to a wide range of fields such as optoelectronics, nonlinear optical effects, or laser engineering for the generation, control, or measurement of light. Commonly used optical devices can be used as the working device 112 and the non-linear device 111, and sometimes dedicated implementations are used. These include optical amplifiers (photon multitube, etc.), absorbers (attenuating materials), reflectors, mirrors, scatterers, collimators, (variable focus) lenses, deflectors, polarizers, polarizing filters, ND filters, polarized beams Splitters (separation), shields, optical waveguides (using photonic crystals, etc.), optical fibers, optical Kerr effect devices, nonlinear optical fibers, optical mixing optical fibers, optical fibers for modulation, optical confinement devices, optical memories, couplers , directional coupler, distributor, mixer/distributor, spectrometer, dispersion-shifted optical fiber, bandpass filter, phase conjugator (by degenerate four-wave mixing, photorefractive effect, etc.), switch by optical control of ferroelectric semiconductor, phase Delay devices, phase correction devices, time inverters, optical switches, encoding by optical masks, etc. may be used alone or in combination. Moreover, it is not limited to this. Under optical control (wavelength conversion/switching/routing), optical node technology, optical cross-connect (OXC), optical add/drop multiplexing (OADM), optical multiplexer/demultiplexer, or optical switch element is used, and the device itself constitutes an optical transmission network or an optical network, and optical signal processing may be performed.

A CCD camera, a photodiode, a mixed photodiode, or a virtual source (also as a wave source) as described herein can also be used for detection. Optical signal processing includes temporal and spatial filters, correlation calculation and matched filter processing, signal extraction, heterodyne and superheterodyne (which can also be used for AD and demodulation by obtaining low-frequency signals), homodyne, etc., and electromagnetic wave detection. utensils are sometimes used.

Examples of nonlinear media in particular include carbon disulfide, sodium vapor, semiconductors such as silicon and gallium arsenide, quantum wells, and organic dyes such as fluorescein and erythrosine. Crystals such as barium titanate also have self-pumping that allows four-wave mixing to occur without supplying pump waves from the outside.

For other waves such as visible light, infrared rays, microwaves, terahertz waves, and radiation, each general-purpose device can be used, but a dedicated device may be realized and used. Not only SAWs, but also devices that have a relationship between a vibration system and an electromagnetic system are useful. Non-linear devices can also be used. Alumina and zirconia composites, solders, and layered cobalt oxides, among others, exhibit nonlinearities in heat conduction. Heat can also act on optical devices and produce nonlinearities, making it possible to consider their applications broadly.

The transducer 110 can be of a contact type or a non-contact type with respect to the object to be measured, and an impedance matching device for each wave may be required as a working device. Matching devices are sometimes used between devices in the measurement space, as well as between devices in an apparatus and in an electric circuit. For example, gel or water is used as a matching material when observing living tissue with ultrasound. In ultrasonic microscopes, samples are usually observed on a stand, but array type and mechanical scan type (elements or element arrays move mechanically through a matching material such as water in a housing) , etc.) have been realized and used. It is also possible to directly observe in situ or in vivo conditions without excision (in vitro). Many ultrasonic microscopes are of a fixed type in which the focal position is determined by a lens or the like, and this method is particularly good when such an element or element array is used. Therefore, mechanical scanning is not limited to lateral and elevational directions, and may also be movable in the propagation direction. An antenna is used for RF waves, and an electrolyte gel and electrodes or a SQUID meter is used for each observation of biological tissue potential and magnetic field. , a miniaturized version is sometimes used (e.g., a microscope). A weak signal may not have non-linearity, and in such cases, non-linearity may be generated artificially or virtually. If the nonlinear signal is too weak to be observed, then enhancement of the nonlinearity is also performed.

Non-linear devices may be integrated with transmitter 121 or transducer 110, or non-linear devices may be assembled and used separately. In this way, non-linear devices can perform non-linear calculations not only on received signals, but also on waves themselves by using non-linear devices at arbitrary positions, including increasing the frequency and bandwidth. can.

In addition, the present invention may be applied to the case of passively observing waves, including the case where the wave source cannot be controlled. Various methods or devices may be used to determine the signal source position, direction of arrival, intensity of the signal source, size and distribution of the signal source, and the present invention may be applied. In some cases, the position of the source and the direction of arrival are also required. At that time, they may be obtained by separating waves and signals, or after obtaining the sources and directions of arrival of signals and waves, waves and signals may be separated, or both may be obtained at the same time. Sometimes. Once the wave source and direction of arrival are determined, the accuracy of receive beamforming improves. The signal is subjected to signal processing such as analog processing or digital processing, and the wave can be time or space, or a filter or spectroscope for those frequencies can be used.

When a transducer receives a wave propagating through a medium including an object to be measured, the transducer used for transmission may also be used for reception (when observing a reflected signal). On the other hand, a different transducer may be used for reception than the one used for transmission. In that case, when the transmitting transducer and the receiving transducer are in close proximity (when observing a reflected signal), or when the transmitting transducer and the receiving transducer are in different positions (for example, when observing a transmitted wave or a refracted wave) In some cases.

Also, the transducer 110 may have a single aperture, or a plurality of transducers 110 may be arranged in an array (one-dimensional array, two-dimensional array, or three-dimensional array) in close proximity. There are also sparse arrays, or ones placed at distant positions, that are used at the same time. There are various shapes of openings (circular, rectangular, flat, concave, convex, etc.), and their directivity is various. Some elements are integrated with openings in a plurality of directions, and some have multidirectional directivity at the same position. In addition to scalar measurements such as potential, pressure, or temperature, there are also those that perform vector measurements such as electromagnetic waves and electric field vectors. Some are polarized. Of course, even for the same wave, the materials and structures of the elements are diverse. In addition, there are various arrangements using them, and for example, there are those with openings facing in multiple directions.

FIG. 39 is a schematic diagram showing an arrangement example of a plurality of transducers. In FIG. 39, (a1) shows a plurality of transducers 110 densely arranged in a one-dimensional array, and (b1) shows a plurality of transducers 110 sparsely present in one dimension. . (a2) shows a plurality of transducers 110 arranged densely in a two-dimensional array, and (b2) shows a plurality of transducers 110 that exist sparsely in a two-dimensional array. (a3) shows a plurality of transducers 110 densely arranged in a three-dimensional array, and (b3) shows a plurality of transducers 110 sparsely present in three dimensions.

The beam may be generated or conditioned analogously using lenses or the like at the aperture of the transducer, but may also be conditioned by the drive signals described above. Also, the imaging apparatus according to this embodiment comprises a mechanical scanning device having at most six degrees of freedom (three translational and three rotational degrees of freedom), wherein the mechanical scanning device comprises at least one transducer 110 or at least one transducer. By mechanically moving the array in at least one direction, the object 1 may be scanned, focused and steered.

On the other hand, when multiple transducers 110 are used, transmitters 121 are provided with a number of channels equal to the number of transducers 110 to generate a number of drive signals equal to the number of transducers 110 to be driven. Alternatively, a group of delay elements can be used to generate multiple drive signals from a limited number of generated signals for desired beamforming (focusing at a desired position or steering in a desired direction). ) may be performed.

Ordinary analog or digital beamformers can also be used. The above-mentioned beamforming (including the case of reception only) may be performed in parallel to improve real-time performance when scanning the measurement target.

Also, at the same time, a plurality of transducers 110 may be driven to simultaneously perform a plurality of beamformings. Alternatively, beamforming may be performed multiple times using different transducers 110 at different times within a time period during which signals in the same phase are allowed to be received, including when the transmitters 121 are switched. The same transducer may be mechanically scanned and beamformed multiple times.

In each beamforming, including the case of mechanical scanning, classical aperture plane synthesis may be performed, either conventional delay-and-summation or delay-multiplication according to the present invention. -and-Multiplication) is performed (either monostatic or multistatic). A plane wave may be generated without focusing at the time of transmission, in which case it is possible to observe a wide area at once in a short time. In doing so, the plane wave may be steered. Waves can be received as plane waves, or they can be dynamically focused (if you steer on transmit, you should also steer on receive). A plane wave is a wave with a narrow band in a direction orthogonal to the direction of propagation, and is effective when broadband.

FIG. 40 is a diagram for explaining the form of waves when using a one-dimensional transducer array. In FIG. 40, (a) shows wave focusing, and a wave beam narrowed down to a focus position determined by delay time setting is formed at the time of transmission or reception. (b) shows wave steering, and a wave beam deflected in a direction determined by setting the delay time is formed at the time of transmission or reception. (c) shows transmission or reception of a plane wave, and a plane wave is formed in a direction determined by setting the delay time. A plane wave is a wave with a narrow band in a direction orthogonal to the direction of propagation, and is effective when broadband.

By passing the waves through at least one of an analog delay device and an analog storage device before being received by the transducer 110, a plurality of waves can be superimposed at each position within the measurement object 1. be. Also, the analog signal obtained after reception by the transducer 110 is passed through at least one of an analog delay device and an analog storage device, or the digitally sampled digital signal obtained after reception is delayed by a digital operation, Alternatively, multiple waves may be superimposed at each location within the measurement object through passage through a digital storage device. A so-called phase aberration correction may be implemented as described above or in conjunction with the beamforming phasing described above. There are various devices. For example, in light, an optical fiber can be a delay line, and an optical confinement device can be a delay device or a storage device.

On the other hand, regarding the object to be measured, there are cases in which a signal strongly affected by a nonlinear phenomenon is observed as a result of propagation in the object to be measured, but conversely, there are cases in which no nonlinear component is obtained. In general, non-linear phenomena are more likely to be observed when the intensity of waves is high, and are less likely to be observed when the intensity is low. The present invention can be practiced in either case. The received signals may be separated into independent signals, such as by performing appropriate signal processing, to implement the present invention.

For signal separation, various wave analog devices (time or space, or their frequency filters, spectroscopes, etc.) may be used, and based on signal processing, analog processing or digital processing ( Decoding processing for the above coding, processing to obtain the centroid of the spectrum through spectrum analysis, processing to obtain the analysis signal to obtain the instantaneous frequency, MIMO, SIMO, MUSIC, or independent signal separation processing, etc.) be. In the passive case, various methods or devices may be used to determine and process the source location, direction of arrival, source strength, source magnitude and distribution, and the implementation of the present invention. The signal source position and direction of arrival may be obtained later. The position of the signal source, the direction of arrival, the strength of the signal source, the size and distribution of the signal source are sometimes obtained simultaneously with beamforming. As will be described in detail later, non-linear processing may be applied to accurately separate signals in situations represented by harmonics or the like. Specifically, after performing processing for increasing the frequency and broadening the band (when the degree is greater than 1) or decreasing the frequency and narrowing the band (when the degree is less than 1) by exponentiation, in the frequency domain , may be performed with high precision. Restoration of the signal after separation is also easy by multiplying the reciprocal of the exponentiation order used.

FIG. 41 is a diagram showing the angle of the beam direction and wave arrival direction and the center of gravity of the spectrum in the spatial domain and the frequency domain in the case of two-dimensional measurement. In FIG. 41, (a) shows the beam direction angles θ1 and θ2 of beam 1 and beam 2 at the point of interest (x,y) in the spatial domain. (b) shows the centroid of the spectra of beam 1 and beam 2 and the instantaneous frequency (fx, fy) of beam 1 in the frequency domain. Beam 1 or beam 2 may correspond to waves of grating lobes or side lobes.
By the way, if the element pitch is coarse in the array type sensor, the signal that causes aliasing in the element array direction (originally digital space) will be received and beamforming will be performed, so the angular spectrum of the received raw signal or after beamforming In the signal, the folded band signal is filtered out. If the element width is reduced and the element pitch is shortened as described above, the band becomes wide in the horizontal direction as can be confirmed by the angular spectrum, and a signal with a wide band in the horizontal direction can be generated by beamforming. need to be processed. These processes are necessary in the case of all beamforming processes. Since beamforming signals can be generated within the signal band of the angular spectrum of the received raw signal, the steering angle achievable with the element array can be ascertained.

Basically, either the beamforming is done analogously to the waves, or if multiple transducers 110 are used, the beamforming (focusing or steering) is done. As described above, beamforming may be performed after signal separation, but signal separation may be performed after beamforming.

Also, when performing aperture plane synthesis processing, it is possible to generate focusing signals for different positions and steering signals for different directions from the same set of received signals (Delay-and-Summation or Delay-and-Summation based on the present invention). -and-Multiplication). It is also possible to implement the present invention on those generated signals. Transmitter 121 and receiver 122 may be integrated or separated.

The non-linear element 124 can be of various types and can be a diode or a transistor for electrical analog signals received at the transducer 110 . In addition, any nonlinear element that applies a nonlinear phenomenon to a signal by a circuit can be used, including those that apply superconductivity. A nonlinear element of a distributed constant system can also be used. An appropriate one is used according to the frequency of the wave (signal). Waves or signals may be appropriately gain adjusted using various amplifiers and attenuators.

Non-linear operations may also be performed using non-linear devices that induce non-linear phenomena directly on the waves before they are received at transducer 110 . For example, with respect to light, the above (i) nonlinear optical element, (ii) light mixing device, (iii) optical parametric effect, (iv) multiphoton transition such as ordinary Raman scattering (spontaneous emission Raman scattering), (v ) non-linear refractive index change, or (vi) electric field dependent refractive index change, etc. can be used. For other waves, nonlinear devices can be used as well. These nonlinear devices may be integrated with the transducer 110, or they may be assembled separately and used. Also, the nonlinear phenomenon itself at the time of conversion from waves to electrical signals in the transducer 110 used for reception may be used.

In any of the above cases, there are cases where analog calculation (non-linear processing) is applied to the wave itself, and analog calculation is applied to the received signal. A non-linear operation may be applied to the signal using a computer, a device similar thereto (FPGA, DSP, etc.).

Regarding the imaging apparatus according to one embodiment of the present invention, when it is called an analog type, it means that the calculation is based on analog processing as described above. digital) or the like, and recorded in a storage medium such as a photograph (analog or digital) or holography, if necessary. Alternatively, the data may be digitized through AD conversion, recorded in a digital data storage medium such as a memory, hard disk, or CD-RW as required, and displayed using a display device.

On the other hand, when it is called a digital type, digital non-linear arithmetic processing includes the case where the analog signal is AD converted after appropriate analog processing (gain adjustment and filtering), and the signal is stored in the recording medium such as memory or hard disk. data is stored in a data storage device (such as the above photographs and digital recording media) as required, and displayed on a display device.

In the above configuration, the received signal may contain nonlinear phenomena in the object to be measured. In such cases, the above analog or digital device can be used to enhance the effect, but the nonlinear component In a received signal that does not contain , it is possible to generate new nonlinear effects, simulate nonlinear effects, or virtually realize them. Further, the nonlinear effect (harmonic component) generated in the measurement object, the nonlinear component (harmonic component) generated in the signal source, and the effect of the nonlinear operation are sometimes separated. Exceptionally, the above devices and signal processing may be used to separate the former two nonlinear effects (nonlinear components), including cases where nonlinear operations are not performed.

In the above explanation of the imaging apparatus, the case of using a transducer for the wave motion to be observed was described. Alternatively, the propagation of shear waves, which predominate as low-frequency oscillatory waves in human tissue, can be observed using the Doppler effect of ultrasound, which is also an oscillatory wave.

It is also possible to optically capture the propagation of sounds such as audible waves and ultrasonic waves. Optical processing generally refers to processing electromagnetic waves, and includes radiation such as X-rays. Ultrasound may also be used to observe audible sound waves. Thermal waves can also be observed by using infrared cameras based on radiation, changes in sound velocity of microwaves, terahertz waves, and ultrasonic waves, volume changes of objects, chemical shifts in nuclear magnetic resonance, or optical fibers. This is due to coherent signal processing or incoherent processing such as image processing. Other waves can be used to observe wave behavior of interest, but are not the only cases where the measurement result is either an analog or digital signal. Therefore, the present invention can also be implemented for those observed waves (signals). In addition to the Doppler effect, it is sometimes interpreted that the physical properties of the medium are modulated by the waves to be observed, which modulates the waves used for sensing. In these, processing for detecting waves subjected to Doppler effect and modulation is useful. In particular, electromagnetic waves can easily observe waves propagating in various directions by applying polarization, and can easily capture structures in various directions. On the other hand, sound waves also allow various measurements based on divergence, as described herein. Radiometric measurements are also important. In addition to temperature distribution measurement, microwaves are used for various remote sensing. For example, scattering and attenuation are measured to measure the distribution of raindrops, moisture, atmospheric pressure, and the like. Even in such a case, it is effective to obtain high spatial resolution by various processes including beamforming described in the specification of the present application, and in particular, the effect of being able to observe a desired position at high speed is obtained. It is possible to obtain the effect of being able to observe an arbitrary surface, area, and space directly and at high speed without relying on image processing after image generation.

In addition, in the above description of the imaging device, non-linear arithmetic devices for electromagnetic waves, vibrations including sound, heat waves, or signals corresponding to them were described, but non-linear effects between different types of physical energy can be enhanced or , simulating a non-linear effect, or virtually realizing a non-linear effect In that case, it is possible to receive waves using devices related to multiple types of waves used at the same time, or to receive signals at different times if they are in phase. As a basis, the invention can also be implemented. That is, the present invention can handle cases in which multiple types of waves are generated simultaneously and cases in which one type of wave is generated independently.

In addition, in vibrations including electromagnetic waves and sound and waves of heat, if the frequencies differ, the dominant behavior differs depending on each measurement object (medium), and the name differs (it may be considered that the type differs). For example, electromagnetic waves include radiation such as microwaves, terahertz waves, and X-rays. As for vibrations, shear waves do not propagate as waves in the megaHz band due to the effects of attenuation in human soft tissue. , ultrasonic waves are dominant, but at low frequencies such as 100 Hz, incompressible features are strong and shear waves are dominant.

The present invention can enhance the nonlinear effect between waves with different behaviors, simulate the nonlinear effect, or virtually realize it. In that case, the invention can be practiced on the basis of signals received at different times, if they are in phase, using devices related to multiple types of waves at the same time to receive the waves. Of course, phenomena such as attenuation, scattering or reflection of those waves have dispersive characteristics and have limitations that must be used appropriately in consideration of the signal-to-noise ratio of the received signal. However, the scope of application of the present invention is very wide, including the ability to generate high frequency components that cannot be physically captured.

It should be noted that when positively observing the nonlinear effect in the object to be measured and when positively performing the nonlinear processing according to the present invention, the two may be used by switching between them, or both may be used simultaneously. Elucidation of non-linear effects within the measurement object may also be performed through computation.

Next, an embodiment in which the present invention is applied to ultrasonic echo signals using the configuration of the imaging apparatus described above will be described. The generation of harmonics in the ultrasound propagation process is represented by multiplication or exponentiation. In particular, chords and difference tones are represented by multiplication of waves with different propagation directions or frequencies (see Non-Patent Document 27), and harmonics are generally represented by powers of waves of the same frequency ( See Non-Patent Document 25). As a physical phenomenon, it tends to occur when the intensity of the wave is high. Wave distortion also has the effect of increasing with propagation for high intensity components, but is more susceptible to attenuation during propagation than the fundamental wave. On the other hand, when the intensity is not so strong, only superposition (sum or difference) is strongly observed as wave interference. See Non-Patent Documents 13 and 30, etc.).

FIG. 42 is a diagram showing in two-dimensional space two deflected beams used for the transverse modulation method. In FIG. 42, the horizontal axis indicates the horizontal position y, and the vertical axis indicates the depth direction position x. Here, as a typical example, two cases of beam forming in one arbitrary direction (direction of angle θ in the figure) and lateral modulation with one arbitrary direction as an axis (X axis) are shown. We confirm the effect of nonlinear operations after receive beamforming for two cases. Note that this calculation can be easily extended to a three-dimensional space, and it can be confirmed that a similar effect can be obtained even in a three-dimensional space. In the following, "λ" is the wavelength corresponding to the center-of-gravity frequency of ultrasound. In addition, the distance x in the depth direction and the distance y in the horizontal direction are the time t for the ultrasonic wave transmitted from the origin to be reflected at a certain point and return to the origin, and the ultrasonic wave propagates at time t/2. represents distance.

<0> Transverse modulation: Superposition of two beams or waves (plane waves, etc.) in the directions of angles θ ₁ and θ ₂ (simultaneous transmission and reception or superposition of each transmission and reception)
The superposition (addition or summation) of the two RF echo signals is given by the following equation.
A(x,y)cos[2π(2/λ)( _xcosθ1 ＋ _ysinθ1 )]
+A'(x,y)cos[2π( ₂ /λ)(xcosθ2+ _ysinθ2 )] (50')

Here, assuming that reflection or scattering is equal and A(x, y)=A'(x, y), coordinates consisting of the X-axis in the direction of the middle of the propagation directions of the two waves and the Y-axis orthogonal thereto At (X, Y), the superposition of the two RF echo signals is given by
A(x,y)cos{2π(2/λ)cos[(1/2)(θ ₂ −θ ₁ )X]}
×cos{2π(2/λ)sin[( _1/2 )(θ2 ₋ θ1)Y]} (50)
Thus, lateral modulation is achieved in the (X, Y) coordinate system. The two waves may be of different frequencies. For example, in <2> and <3> below, non-linear processing is performed. In three-dimensional space, there are two directions for lateral modulation, and therefore at least three intersecting beams must be generated (see Non-Patent Documents 13 and 30).

<1> Power calculation of one beam or one wave in the θ direction An RF echo signal is expressed by the following equation.
A(x,y)cos[2π(2/λ)(xcosθ＋ysinθ)]
In this case, for example, the square is represented by the following equation (51).
(1/2) ^A2 (x,y)×{1+cos[2π(2 2/λ)(xcosθ+ysinθ)]} (51)
Thus, the second harmonic component is generated at the same time as the DC component, and the basebanded signal is obtained at the same time (the envelope signal is also obtained directly). The calculated squared echo signal has a wider bandwidth than the spectrum of the fundamental wave due to the product of different frequencies within the band, resulting in shorter pulse lengths and beam widths and higher spatial resolution.

As _a more intelligible example, for example, when the RF echo signal at the position in the depth direction x has frequencies f1 and f2, the _squared echo signal is represented by the following equation by squaring.
_eI (x;f1, _f2 ) ₂ ⁼ _{eII ( x} ; _0,2f1,2f2 , _{f1 +} f2,f1 _- f2)
Thus, the squared echo signal will have DC (frequency 0) and signal components at frequencies 2f ₁ , 2f ₂ , f ₁ +f ₂ and f ₁ -f ₂ .

That is, the signals generated by the exponentiation operation have a plurality of different frequency signal components compared to the wave received without the nonlinear operation when the wave has signal components of different frequencies. , and the harmonics are higher in frequency, higher in spatial resolution, or lower in side lobes, or It is a signal that has obtained at least one effect of increasing the contrast, and the signal (base banded signal) generated in the band including DC is a signal obtained by substantially quadrature detection of harmonics. Waves can be imaged based on at least one.

Higher-order exponentiation yields a signal component with a frequency n times as high as that of the fundamental wave by the n-th power, and further enhances the spatial resolution. Strictly speaking, the basebanded signal differs from the quadrature-detected version of its second harmonic (ordinary baseband signal). A high-resolution image can be easily obtained compared to the echo image of Note that the DC component generated by the nonlinear operation is obtained from the strength of the high frequency, low frequency, harmonics, etc. generated at the same time, and the DC component contained in the basebanded signal is basically excluded. Sometimes the computation is simplified to remove all DC in the basebanded signal. This processing may allow deeper imaging without depth-dependent brightness adjustment compared to direct current.

According to the double-angle or arc-angle theorem, harmonic signals and low-frequency signals are expressed in various forms (four arithmetic operations of sine waves and cosine waves), and when necessary, digital Hilbert transform (see Non-Patent Document 13 (see ). Measured harmonics can also be used. These are calculated nonlinear signals at each position for arbitrary intensity waves, and are new harmonic or low frequency components, unlike nonlinear components that are physically accumulated and subject to attenuation during the propagation process. It realizes frequency imaging.

<2> Power Calculation of Laterally Modulated Echo Signal For example, the square of Equation (50) is expressed by Equation (52) below.
A(x,y) ² ×cos ² {2π(2/λ)cos[(1/2)(θ ₂ −θ ₁ )X]}
×cos ² {2π(2/λ)sin[(1/2)(θ ₂ −θ ₁ )Y]}
＝A(x, y) ² ×[1＋cos{2π(2・2/λ)cos[(1/2)(θ ₂ −θ ₁ )X]}
＋cos{2π(2・2/λ)sin[(1/2)(θ ₂ −θ ₁ )Y]}
＋cos{2π(2・2/λ)cos[(1/2)(θ ₂ −θ ₁ )X]}
×cos{2π(2 2/λ)sin[( _1/2 )(θ2 ₋ θ1)Y]}] (52)
Thus, two signals are obtained: DC (the basebanded signal described above), a second harmonic detected in a different direction, and a laterally modulated signal of the second harmonic. Similar to <1>, high resolution is also performed. Basebanded signals and other high-order harmonic signals can also be calculated in the same manner as <1>.

As _a simple example, the cross _- echo signals at location (x,y) are e1((x,y);( _f0 ,f1)) and e2 ₍ (x,y);( _f0 , f ₂ )) and symmetric in the y-direction, the squared signal of the superposition signal is given by:
[ _e1 ((x,y);( _f0 ,f1)) ^＋ e2 ₍ (x,y); _{( f0} ,f2))] ₂
＝ _e1 ((x,y);( _f0 ,f1)) ₂ ^＋ _2e1 ((x,y);( _f0 , _f1 ))e2 ₍ (x,y);( _f0 , f ₂ )) + e ₂ ((x,y);(f ₀ ,f ₂ )) ²
＝e ₁ '((x,y);(0,0),(2f ₀ ,2f ₁ ))+e ₁₂ '((x,y);(2f ₀ ,0),(0,2f ₁ ), ( _0,2f2 ))
₊ e2'((x,y);(0,0), _{( 2f0,2f2} ))
Thus, the squared signals of the superposed signals are at frequencies ( _0,0 ), ( _2f0,2f1 ), ( _2f0,2f2 ), ( _2f0,0 ), ( _{0,2f1 )} , ( 0,2f ₂ ).

That is, the signals generated by the exponentiation operation are the harmonic signals of each linearly superimposed signal (the signal corresponding to the crossed waves) and the basebanded signal (of a band containing DC in at least one direction). signal), and in the case where the wave has signal components of different frequencies, it is broadened in the direction of the wave that has a plurality of different frequency signal components compared to the wave that is received when non-linear calculation is not performed. and the harmonics are at least one of higher frequency, higher spatial resolution, lower sidelobes, or higher contrast than the corresponding waves received without the nonlinear operation. The basebanded signal is a signal obtained by quadrature detection or substantially quadrature detection of harmonics in each direction or multiple directions, and based on at least one of these signals, the wave can be imaged. can. If the intersecting waves or beams have different frequencies or are not symmetrical with respect to the coordinate axes, the same process can produce chords and difference tones in multi-dimensional space. used for They work even in situations where other parameters are different.

In three-dimensional space, lateral modulation must produce at least three crossing beams, as described above, but in this case the resulting basebanded signal is the harmonics of each beam as well as approximately In addition to quadrature-detected signals (neighboring signals including DC), quadrature-detected signals in only one arbitrary direction or in two arbitrary directions can be obtained. That is, since the polarities of the frequencies in the symmetrical direction are opposite to the symmetrical axis, the sum of the frequencies becomes zero. All waves or beams may be generated symmetrically about the coordinate axes, but this is not the case. Also, the frequencies and other parameters may be different.

<3> Multiplication of two waves of laterally modulated echo signals For example, since the two waves in the equation (50′) can be treated separately, when considering the product, the propagation direction is x Assuming two directions symmetrical with respect to the axis, θ ₁ =−θ ₂ , and the multiplication (product) of the two RF echo signals is expressed by the following equation (53).
A(x,y)cos[2π(2/λ)( _xcosθ1 ＋ _ysinθ1 )]
×A'(x,y)cos[2π(2/λ)(xcosθ ₁ -ysinθ ₁ )]
＝A(x, y) A'(x, y)×{cos[2π(2・2/λ) _cosθ1x ]
+cos[2π(2 2/λ)sinθ ₁ y]} (53)
As a result, two signals of second harmonics detected in different directions are obtained. These are the signal components also obtained in equation (52).

As _a simple example, the cross _- echo signals at location (x,y) are e1((x,y);( _f0 ,f1)) and e2 ₍ (x,y);( _f0 , f ₂ )) and symmetric in the y-direction, the multiplication of the signals is given by
e1((x,y);( _f0 , _f1 )) _× e2 ₍ (x,y); _{( f0} ,f2))
＝e ₁₂ '((x,y);(2f ₀ ,0),(0,2f ₁ ),(0,2f ₂ ))
Thus, the multiplication of signals can be seen to have signal components at frequencies (2f ₀ ,0), (0,2f ₁ ), (0, 2f ₂ ).

That is, the signal generated by the multiplication operation is a basebanded signal (a signal in a band containing DC in at least one direction) for each linearly superimposed signal (signal corresponding to the crossed waves), If the wave has signal components of different frequencies, it is broadened in the direction of the wave that has the different frequency signal components compared to the wave that is received when non-linear calculation is not performed, and is basebanded. The signal has at least one of a higher frequency, a higher spatial resolution, a lower sidelobe, or a higher contrast than the corresponding waves received without the nonlinear operation. It is a signal obtained by quadrature detection of harmonic signals in each direction or multiple directions, and wave motion can be imaged based on at least one of these signals. If the intersecting waves or beams have different frequencies or are not symmetrical with respect to the coordinate axes, the same process can produce chords and difference tones in multi-dimensional space. used for They work even in situations where other parameters are different.

In three-dimensional space, lateral modulation must produce at least three crossing beams, as described above, but in this case the resulting basebanded signal can be in any one direction only or in any two directions. A direction quadrature detected signal is obtained. That is, since the polarities of the frequencies in the symmetrical direction are opposite to the symmetrical axis, the sum of the frequencies becomes zero. All waves or beams may be generated symmetrically about the coordinate axes, but this is not the case. Also, the frequencies and other parameters may be different.

In addition to the case where the propagation direction and steering angle of each beam and wave are different like the above cross beams, another parameter is different, for example, frequency, carrier frequency, pulse shape, beam shape, frequency seen in each direction , carrier frequency, or bandwidth may be different. Also, in each dimension, unlike creating two (or sometimes three) intersecting waves or beams in 2D or 3D in transverse modulation, , more waves or beams may be used. In particular, when plane waves, cylindrical waves, and spherical waves are transmitted, high-speed transmission and reception are possible. In addition, even when focusing beams are used, it is possible to perform fast beamforming using fast Fourier transforms on the superimposed received signals, especially when multiple beams are transmitted simultaneously. (As mentioned above, interpolation approximation may be performed in wavenumber matching). In order to stabilize the nonlinear processing, it is also effective to transmit and receive multiple times with the same parameters and superimpose them (additional averaging). In addition, the same processing as described above is also possible for signals received when so-called pulse inversion transmission is performed. It is possible to perform the same processing on , or to perform the same processing on them before superimposition. When they are superimposed (that is, added), you get harmonics with frequencies that are even multiples of the frequency of the fundamental, but if you subtract instead of add, you get odd harmonics. Their use in imaging is also important (a mere subtraction of the pulse inversion received signal yields primarily the 3rd harmonic). Filtering (analog or digital ), or signal processing (analog or digital) with various superpositions and fundamentals to separate the harmonics. Also, instead of pulse inversion, signals having a phase difference other than 180° may be transmitted, and the present invention can also be applied in such cases. That is, even in beams or waves in which at least one of the parameters is different, the same nonlinear effect can be obtained in a superimposed state, a separated state, a non-superimposed state, etc., and can be effectively used. Sometimes. Through theory or calculation, it can be understood that waves and beams generated by not only linear effects but also nonlinear effects can be designed (parameters of waves and beams such as direction of propagation) and controlled.

Harmonic signals, chords, difference tones, overtones, etc. generated by these non-linear operations have the above characteristics and improve the image quality of echo imaging. There is also no attenuation effect that occurs in normal harmonic imaging. The present invention is also useful for interpreting physically induced nonlinear signals that produce a nonlinear component virtually at each location. Moreover, the present invention is effective even when the signal is weak and cannot be observed. Furthermore, in displacement measurement, higher frequencies are welcomed, and the phase rotation becomes faster, so it is expected that high-precision measurement will be possible. , the noise tends to increase when the high spatial resolution is measured as it is.

In such a case, regularization (see, for example, Non-Patent Document 18), the weighted least-squares method through the above statistical evaluation, averaging, and the like are effective. For example, the two signals of the second harmonic detected in different one directions obtained in <2> and <3> can be used for displacement measurement in each direction using the usual one-way displacement measurement method. can be done. In order to measure the displacement or displacement vector occurring between different time phases, for a signal having a carrier frequency in only one arbitrary direction, at each point of interest, the instantaneous phase change occurring between the time phases is measured as the instantaneous frequency and the centroid frequency. Or, the displacement in that direction can be measured by dividing by the nominal frequency, etc., and a displacement vector can be synthesized based on the measurements in the different directions. In the past, digital A demodulation method was devised and reported (calculating products and conjugate products of analytic signals: see Non-Patent Document 30, etc.). According to the present invention, transversely modulated echo signals can be demodulated with a small amount of memory and computation in each stage, and the signals obtained are harmonic signals. In addition, noise can be averaged after performing non-linear processing after the received raw signal or the received signal when multiple waves of the same wave can be acquired under the same conditions. It is effective to reduce it by, for example, performing integration processing after performing non-linear processing after reception. Also, instead of exponentiation and multiplication, square norms and inner products can be calculated, in which case the spatial resolution is determined by the signal length during the calculation. These methods may also be effective in imaging other than displacement measurement.

The digital demodulation method described in Non-Patent Document 30 invented by the inventor of the present application specifically derives the phase determined only by the displacement component in each direction, and obtains the displacement component in each direction. For example, when measuring a two-dimensional displacement vector (dx, dy), the instantaneous phase difference between two different time phases at a point in a two-dimensional region of interest is generated by two intersecting beams or waves. expressed as the phases of the complex autocorrelation signals exp[j(fxdx+fydy)] and exp[j(fxdx−fydy)] using the spectra of two independent single quadrants that are By calculating the product and conjugate product of , we obtain exp[j(2fxdx)] and exp[j(2fydy)], and the instantaneous phase differences 2fxdx and 2fydy in each direction at the instantaneous frequencies 2fx and 2fy in each direction. By dividing, an unknown displacement vector (dx, dy) is obtained. Also, when measuring a three-dimensional displacement vector (dx, dy, dz), exp[j(fxdx+fydy+fzdz)], exp[j( fxdx + fydy - fzdz)], exp [j (fxdx - fydy + fzdz)], exp [j (fxdx - fydy - fzdz)], or at least three of them, can be easily obtained in the same way . This digital demodulation method or the nonlinear calculation of <2> or <3> has a carrier frequency in each direction of an arbitrary wave symmetry axis crossing an arbitrary direction and a direction orthogonal thereto. Since waves (waves detected in one or two directions) are generated, avoid places where there are obstacles or shields related to those waves and cross the waves in that way behind them. can generate a wave with a carrier frequency in an arbitrary direction behind obstacles and shields, etc., which cannot be generated directly through obstacles and shields. It is possible to perform imaging and displacement measurement of the rear of the For example, when a wave with a carrier frequency is generated in the depth direction and lateral direction through an obstacle or shield, it is equivalent to realizing a situation in which the obstacle or shield is seen through from the front direction. Yes, and at that time, the movement of the target in any direction can also be measured. This imaging and displacement measurement can be performed not only from the front direction of obstacles and shields, but also from any direction. In those cases, at least one mirror may be used to generate reflected waves to observe behind obstacles, shields, and the like. For example, a wave steered from the front direction of an obstacle or shield is generated, reflected by a mirror in the steering direction, and the wave crosses behind the obstacle or shield. A wave source may exist in a direction other than the frontal direction. Various combinations of steering angle and carrier frequency can be used, but given the need to obtain the effects of complex products and complex conjugation, the method of solving the system of equations is not constrained with respect to the combination and is computationally less complex.
In these digital demodulation methods or nonlinear calculations <2> or <3>, in order to generate a frequency that is twice the instantaneous frequency in each direction, the bandwidth is widened in advance based on the Nyquist theorem. It is necessary to perform beamforming on the spectrum, interpolate the number of beams in space, or widen the spectrum (data interpolation) by zero-filling the spectrum other than the signal spectrum in the frequency domain. In other words, it is possible to prevent the folding phenomenon from occurring. These processes are also a kind of signal separation. Alternatively, if aliasing occurs in these digital demodulation methods or the nonlinear calculation of <2> or <3>, the Nyquist theorem can be applied to the signal before processing without widening the band. In the satisfied situation, the same processing for each of them is applied, in which case the instantaneous phase differences in each direction calculated are 2fxdx, 2fydy, 2fzdz, whereas the instantaneous frequencies fx, fy in each direction. , fz in the original signals, multiply them by two and divide their instantaneous phase difference. It does not require processing to widen the bandwidth, has a small amount of calculation, is fast, and has the effect of requiring a small amount of memory. Note that the sum of the instantaneous frequencies in each direction estimated for each wave or each beam may be used instead of twice the instantaneous frequency. In those cases, instead of the instantaneous frequency, the centroid of the spectrum of the original signal (centroid frequency or center frequency) may be obtained and doubled, or similarly summed and used. Alternatively, a value twice the nominal frequency or a pre-determined typical value may be used without calculation. Further, as shown in the schematic diagrams of the spectrum distributions in FIGS. 43 and 44, in a situation where the folding phenomenon occurs, the frequency region of the frequency coordinates where the folding phenomenon occurs is shifted to the positive or negative half-band frequency region. In other words, the doubled frequency can also be calculated directly from the centroid of the spectrum (it can always be calculated because the Nyquist theorem is satisfied for the unprocessed signal). The spectral distribution may be computed for echo signals within a region of interest or region, or may be computed for echo signals in a local region containing each interest point, with attention to points of interest. FIG. 43 shows an example of processing by rereading the frequency coordinate axis when folding occurs in the two-dimensional spectrum of the band 2A (-A to A) in the depth direction due to demodulation during two-dimensional horizontal modulation. showing. FIG. 44 shows an example of processing by rereading the frequency coordinate axis when folding occurs in the two-dimensional spectrum of the horizontal band 2B (-B to B) due to demodulation during two-dimensional horizontal modulation. is shown. Although there are multiple combinations of independent analytic signals, FIG. 43 shows two analytic signals whose instantaneous (center of gravity) frequencies are (fx, fy) and (fx, -fy). In this case, the center-of-gravity frequency in the depth direction can be calculated as 2fx. FIG. 44 shows two analytic signals whose instantaneous (centroid) frequencies are (fx,fy) and (fx,-fy). It shows an example that can be calculated as Folding phenomenon may occur simultaneously in the depth direction and the lateral direction. Similarly, in the three-dimensional case, the frequency domain of the frequency coordinates in which the aliasing phenomenon occurs can be read as the frequency domain of the positive or negative half-band, and the double frequency can be directly calculated from the centroid of the spectrum. In addition, in the same situation where the folding phenomenon occurs, the frequency domain of the frequency coordinates where the folding phenomenon occurs is similarly read as the frequency domain of the positive or negative half-band, and then the frequency domain of the half-band of the zero spectrum is added. It is possible to widen the band so that it satisfies the Nyquist theorem, and obtain the analytic signal by inverse Fourier transform (At this timing, it is also possible to calculate changes in instantaneous frequency and instantaneous phase by widening the band and interpolating the signal. , the amount of calculation can be reduced compared to band widening or interpolation in advance, but the amount of calculation is large compared to those processes without band broadening). Of course, if folding does not occur in these digital demodulation methods or the nonlinear calculation of <2> or <3>, the instant of double or sum that can be obtained directly without rereading the frequency coordinates or broadening the band The frequency or centroid frequency can be used to divide their instantaneous phase difference. Alternatively, a value that is twice the nominal frequency or a pre-determined typical value may be used without calculation. Although there is no problem in the range up to the rereading of the frequency coordinates, if the band is widened by zero padding or interpolation processing of the spectrum, not only the calculation amount will be enormous, but also the accuracy will be lowered. is useful in these respects as well. Regularization processing may be performed as necessary when the band is broadened and when the band is not broadened.

Above, several examples of applying the present invention to ultrasound imaging or ultrasound measurement have been presented. If the band of the signal component calculated and generated by the present invention overlaps the band of another signal, the two cannot be separated in the frequency domain. In that case, the pulse inversion method or the separation of multidimensional terms is used, but the inventor of the present application has reported in the past that the spectrum in the superimposed state is handled or divided and processed ( See Non-Patent Document 30). In the present invention, as another method for accurately separating waves, including the case where spectra are superimposed, in order to obtain the effect of separating them in the frequency space, power In the situation where multiplication is applied and broadbanded and represented as harmonics, they may separate in frequency space. Conversely, a high frequency signal may be reduced in frequency, or a broadband signal may be reduced in frequency to be displayed or handled. High-frequency and wideband (when the order is greater than 1) or low-frequency and narrowband (when the order is less than 1) processing by exponentiation, and then performed with high accuracy in the frequency domain. sometimes be The propagation direction of the wave is determined by the centroid of the spectrum of the generated harmonics (the direction at the local level, i.e., with spatial resolution, or macroscopically with low spatial resolution, or without), or from the analytic signal to the instantaneous frequency ( (with spatial resolution) can be obtained and calculated. Using the order of the implemented exponentiation, other wave parameters such as the frequency and bandwidth of the original wave can be obtained by back calculation, and can be restored in the separated state (the restoration of the signal after separation is also used for the power order can be easily raised to the reciprocal power of ). Including such a case, the signal source position, arrival direction, signal source intensity, signal source size and distribution can be measured with high accuracy. Also, when generating a signal with a frequency higher than that of the original signal, it is necessary to widen the calculable bandwidth in advance before performing the calculation. For this reason, spectral zero padding is effective without approximation (see Non-Patent Document 30), but sampling intervals may be shortened based on interpolation approximation directly in space-time.

Recently, it has become possible to simulate nonlinear propagation at low cost. Therefore, it is also possible to generate a nonlinear signal by applying the nonlinear calculation of the present invention and such a simulation technique to a non-beamforming signal (such as a plane wave) or an echo signal for aperture synthesis. Based on these, it is also possible to analyze (inversely analyze) actually measured nonlinear signals based on an inverse problem approach and apply them to tissue diagnosis.

For example, in the case of ultrasound, it is applied to diagnoses by evaluating sound velocity, bulk modulus, acoustic impedance, reflection, Rayleigh scattering, backscattering, multiple scattering, or attenuation, etc. as biological tissue properties. Sometimes. Inverse analysis of related phenomena and physical properties is also effective for other waves (Mie scattering in light, scattering in radiation, Compton scattering, etc.).

In addition, in the treatment by heating or warming, it is necessary to clarify the heat receiving characteristics of the target (for example, the characteristics of the sound pressure of high-intensity ultrasound, the effect of contrast agents, etc.) and the characteristics of temperature rise. Calculations including nonlinear calculations are effective even in such cases. Also, in therapy, it is useful to evaluate and apply the effects based on the imaging of nonlinear effects according to the present invention. Alternatively, echo imaging and tissue displacement measurement can be performed using the present invention on received signals that have undergone physical nonlinear effects, separated basebanded signals, and multiple harmonics.

In addition to ultrasonic waves, the present invention applies nonlinear operations such as multiplication and exponentiation to coherent signals of electromagnetic waves, light, radiation, mechanical vibrations, sound waves other than ultrasonic waves, and arbitrary waves such as heat waves. The present invention relates to an imaging apparatus that increases the frequency, bandwidth, or contrast of a signal by applying . Harmonic signals can be enhanced, simulated, or newly generated according to the present invention. Furthermore, harmonic signals can be realized virtually.

In addition, it is possible to simultaneously obtain a baseband signal and a detection signal in an arbitrary direction of a harmonic signal with a smaller amount of calculation than in normal detection processing. As a result, for example, higher frequency and wider band, higher contrast, or suppression of side lobes can be achieved, enabling nonlinear imaging with a high SN ratio. In addition, it becomes possible to easily measure a displacement vector with a small amount of calculation using a normal unidirectional displacement measurement method. From the viewpoint of generating chords, difference tones, overtones, etc., it is possible to obtain high-frequency and low-frequency signals, including cases where the frequencies of waves and beams, carrier frequencies, steering directions, or propagation directions are different. and they are sometimes used effectively for imaging and metrology. Through theory or computation, it is also possible to design waves and beams (parameters of waves and beams such as direction of propagation) generated by not only linear effects but also nonlinear effects, and to control them.

On the other hand, in the field of image measurement, various types of detection (including those that simply take the absolute value of the waveform) are applied to the coherent signal to make it incoherent (the result is an image). Observations are also well known. Cross-correlation processing, optical flow, or a method based on the SAD (Sum and Difference) method may be used. Further, even if the present invention is applied to an incoherent signal, it is possible to widen the band (improve the resolution). The above high-resolution detection signal obtained by the present invention can also be used. The situation in which the data has become denser through the increase in bandwidth is suitable for such processing, and the accuracy of motion measurement is also improved. Note that the above method can also be applied to coherent signals, and the widening of the band is effective in improving accuracy. That is, the present invention can be applied to arbitrary coherent and incoherent signals.

Among others, according to the present invention, heating, heating, cooling, freezing, welding, repair, cancer in medicine of any object performed using waves (laser, ultrasound, or focused high-intensity ultrasound, etc.) In the heating of lesions, etc., cryotherapy, or cleaning of arbitrary objects (glasses, etc.), their effects can be enhanced through nonlinear phenomena, high resolution, and prediction of their effects (for example, strong ultra It is possible to improve the effect through the exponentiation effect when heating using sound waves, the effect of not only addition but also multiplication when enhancing the effect using cross beams, etc.).

In heat treatment using high-intensity ultrasonic waves, harmonics are generated by the nonlinear effect of tissue, and because of the high frequency, the absorption effect as heat energy is strong, so heat generation in tissues can be easily understood and predicted. It is also possible to In the same respect, transmitting high-frequency signals, transmitting broadband signals, transmitting harmonics, generating superimposed beams, or generating cross beams are effective for treatment. Yes, again, it is easy to understand and predict. Specifically, the sound pressure shape and point spread function can be estimated on the basis of simulating the sound field or, in systems where the received signal can be obtained, evaluating the autocorrelation function, directly It is also effective to perform non-linear calculation on the fundamental wave signal, if it is possible to evaluate the harmonic signal. The same is true for other waves.

In addition, the present invention can obtain a nonlinear effect under physical conditions where the nonlinear effect cannot be obtained physically (for example, when the intensity cannot be increased for the object to be measured, or when a high intensity cannot be obtained due to high frequency). is also effective. Conversely, the present invention can also be practiced under conditions in which contrast agents such as microbubbles are used to enhance nonlinear effects, for example, during ultrasound echo imaging, displacement measurement, or therapy. It is suitable for measurement and imaging of blood in blood vessels and heart chambers, although the tissue may be targeted while permeated into the tissue. That is, the present invention can enhance, simulate, or generate nonlinear effects. Furthermore, the present invention can also virtually realize non-linear effects. As described above, it is also possible to evaluate non-linear effects. Contrast agents may also be used to enhance the effect of heat therapy. The same is true for other waves.

In addition, when generating a high-frequency signal that cannot be realized with a single signal source, imaging with higher resolution and Doppler measurement with higher accuracy are possible. Generally, the influence of attenuation is strong in high-frequency components. For example, with a microscope that is susceptible to the influence of attenuation, it is desirable to be able to observe as deep as possible at a high frequency. For example, if a plurality of ultrasonic transducers of 100 MHz are used, it is possible to generate ultrasonic waves that are physically twice as high as the number of ultrasonic transducers, realizing a high frequency that cannot be generated by ordinary transducers. It can also simply generate high-frequency signals (chords). According to the present invention, such high frequencies can also be realized by computation. Therefore, according to the present invention, even high-frequency waves and signals that cannot be physically realized can be generated. Similarly, low-frequency imaging and measurements using low-frequency signals (eg, difference tone) can be realized. It is also possible to generate low frequency signals that are physically impossible with a single signal source. These signals can also be implemented theoretically or on a computational basis to control the waves generated.

In the following, experimental data, simulation results, and materials such as photographs will be described in order to prove the effects of the present invention. These demonstrate the effectiveness of the present invention by conducting ultrasonic echo imaging and measurement imaging through ultrasonic simulations and agar phantom experiments. The present invention can be applied to arbitrary signals other than the ultrasonic echo method (familiar ones such as lasers, light waves, OCT, electricity, magnetic field signals, radiation such as X-rays, heat waves, etc.) and between different signals. is. It has applications in raw coherent signals or incoherent signals after signal processing.

For the echo data for aperture plane synthesis of the agar phantom disclosed in Non-Patent Document 30 (linear array type probe, 7.5 MHz), beam forming in the front direction and lateral modulation of 3.5 MHz in the lateral direction The above <1> to <3> were performed using the respective echo signals obtained when performing the above.

FIG. 45 is a diagram illustrating changes in the spectrum of echo signals according to one embodiment of the present invention. In FIG. 45, the horizontal axis indicates the horizontal frequency [MHz], and the vertical axis indicates the depth direction frequency [MHz]. In FIG. 45, (a1) and (a2) respectively show the spectrum of the original echo signal and the spectrum of the square of the echo signal without steering. (b1), (b2), and (b3) represent the spectrum of the original echo signal, the squared spectrum of the unidirectional steering echo signal, and the squared spectrum of the transversely modulated echo signal during transverse modulation. each shown. (c) shows the spectrum of the cross-steering beam echo signal product. From FIG. 45, the spectrum of each signal derived by the above theory can be confirmed. In any echo signal, the squaring or multiplication results in a second harmonic spectrum, which has a wider bandwidth than the original spectrum.

Figures 46A-46C are diagrams illustrating changes in the autocorrelation function of echo signals according to one embodiment of the present invention. Here, the horizontal axis indicates the horizontal position [mm], and the vertical axis indicates the normalized autocorrelation function. FIG. 46A shows a comparison of the autocorrelation function of the original echo signal and the autocorrelation function of the second harmonic by squaring the echo signal for the non-steering case. FIG. 46B shows a comparison between the autocorrelation function of the original transversely modulated echo signal and the autocorrelation function of the squared second harmonic of the transversely modulated echo signal during transverse modulation. FIG. 46C shows the autocorrelation functions of the lateral and depth components for the product of the cross-beam echo signal and the square of the lateral modulated echo signal. Based on the autocorrelation function, the sound pressure and the lateral profile of the point spread function can be evaluated (for a depth of 19.1 mm in the center of the region of interest). Although omitted here, if a two-dimensional autocorrelation function is obtained for this two-dimensional echo signal, the two-dimensional distribution of the sound pressure and the point spread function can be estimated. I would like to find a function.

47-49 are diagrams illustrating changes in B-mode echo images according to one embodiment of the present invention. These echo images have a depth of 10.0 mm to 28.1 mm and a width of 20.7 mm. In the agar phantom, a columnar (diameter 10 mm) clog having a shear modulus about 3.29 times higher than that of the surroundings exists around a depth of 19 mm.

47 to 49, (a1), (a2), and (a3) are the echo image based on the original echo signal, the echo image based on the basebanded signal, and the echo signal in the case of no steering. Echo images based on second harmonic by squaring are shown respectively. If there are two left and right images, the left image is by envelope detection and the right image is by square law detection.

(b1), (b2), (b3), (b4), and (b5) are an echo image based on the original laterally modulated echo signal, an echo image based on the basebanded signal, a lateral An echo image based on the second harmonic by squaring the directionally modulated echo signal, an echo image based on the transverse component of the second harmonic by squaring the transversely modulated echo signal, and a second harmonic by squaring the transversely modulated echo signal. Echo images based on the depth direction component of the second harmonic are shown, respectively.

Also, (c1) and (c2) show an echo image based on the lateral component and an echo image based on the depth component, respectively, for the second harmonic generated by the product of cross beam echo signals. Regarding the case where multiple waves exist, the detection of the superposition of coherent signals has also been reported in the past, but here, the result of superposition of each detected signal is shown.

It can be confirmed from FIGS. 46A to 46C and FIGS. 47 to 49 that the spatial resolution has increased in response to the widening of the spectrum shown in FIG. Here we have not cut the direct current of the basebanded data. If the direct current or, if necessary, the spectrum of extremely low frequencies in the depth direction and the horizontal direction are removed by filtering, etc., high or low luminance lines (vertical stripes) running in the vertical direction can be completely removed (Fig. are omitted). It can be confirmed from FIGS. 46A to 46C that the side lobes are lowered. Correspondingly, from FIGS. 47 to 49, the effect of increasing the contrast can also be confirmed (focus on strong scatterers and the like). Since no attenuation correction was applied to the original echo signal, the image resulting from the processed signal has less signal intensity at depth than the image of the original signal as a result of the increased contrast due to the uncorrected signal. Extremely low compared to that of the position.

In imaging using the original signal, attenuation correction in the so-called wave propagation process is applied to the coherent signal or the incoherent signal after detection. The coherent signal is then subjected to nonlinear processing, or the coherent or incoherent signal is corrected after nonlinear processing. The correction process itself may be performed primarily on a signal strength basis, before or after receive beamforming, or after imaging, as with normal correction. Corrections may be applied according to Lambert's law.

In that case, the average attenuation coefficient may be used, but the attenuation coefficient at each position on the path of the wave or beam is calculated by signal processing or inverse analysis in the calculation unit 130, and the correction can be performed with high accuracy. sometimes implemented. That is, it may be done adaptively or automatically. Alternatively, the operator may adjust the intensity within a predetermined range at each depth via the control unit 133 while viewing the generated image. Selectable patterns may be prepared depending on the object to be measured.

Gain adjustment is performed by the receiver 122, the amplifier or attenuator in the filter/gain adjustment unit 123 or 125, the amplifier or attenuator in the reception beamformer 129, digital processing, or analog processing or digital processing in the calculation unit 130. will be At transmitter 121, the intensity of the transmitted beam or wave may be adjusted. Also, an amplifier or attenuator may be used as the working device 112 to adjust the intensity of the wave itself. Note that the contrast agent 1a greatly influences these determinations.

In the square calculation (<2>) of the transversely modulated echo signal in the above experiment, the transversely detected spectrum and the second harmonic of the two signals of the second harmonic detected in a different direction Since the waves overlapped, the inventor split the spectrum by eye. The result and the result of multiplication (<3>) of the two waves of the laterally modulated echo signal were compared in terms of the autocorrelation function (see FIG. 46C). There was no difference.

In addition to these experiments, we performed displacement vector measurements, strain tensor measurements, and shear modulus reconstructions using the multidimensional autocorrelation method. FIG. 50 shows the result of displacement measurement in each direction using two of the second harmonic signals detected only in one different direction in <3>.

FIG. 50 shows images of the displacement vector, strain tensor, and relative shear modulus measured in an agar phantom according to one embodiment of the present invention. Also shown in part of FIG. 50 are the estimated mean and variance (in parenthesis) in the center of the filling. Compared to the result of applying digital demodulation to the same laterally modulated echo data, there was a tendency for noise to increase (variation in distortion in the lateral direction (y) ranged from 3.08×10 ⁻³ to 9.52×10 ⁻³ ), the spatial resolution is doubled, and regularization on the shear modulus reconstruction results in improved accuracy (from 3.37 to 3.23).

Although the results are omitted here, as described in paragraph 0655, many waves and beams are generated and non-linear processing is performed under the superimposed state, and non-linear processing is performed in the non-superimposed state. may be superimposed after applying Also, as described elsewhere, non-linear processing may be applied to raw received signals that have not undergone beamforming (such as transmit beamforming only or aperture synthesis). Multiple waves or beams generated under the same wave or beamforming parameters may be processed, but waves or beams generated under different parameters may also be processed.

但し、Ｈｐ(ωｘ,ωｙ,ωｚ)とＧ(ωｘ,ωｙ,ωｚ)の各々は、処理対象の信号と目標の信号のスペクトルである。
のノルムに重み付けした

但し、ＰＷpn(ωx,ωy,ωz)とＰＷps(ωx,ωy,ωz)は、各々、ノイズと信号のパワースペクトラムである。ｑは任意の正値である。
を用いて処理すれば良い。上記記載の線形モデルにおける他の超解像においても、同様に、ウィーナーフィルタを応用して、雑音の増幅を抑えることが可能である。ウィーナーフィルタを用いずに、（ＡＡ１）のノルムそのものを施す場合に、信号スペクトルＨｐ(ωｘ,ωｙ,ωｚ)のノルム（最大ノルムやL₂ノルム、L₁ノルム等）のε（＜1）倍以上のスペクトルを持つ周波数のスペクトルのみの処理が行われる（他の周波数のスペクトルは零にする）こともある。これらにおいて、変則的に、（ＡＡ１）のノルムではなく、（ＡＡ１）そのものを用いることもあり、位相まで合わせることもある。その場合に、Ｇｐ(ωｘ,ωｙ,ωｚ)は、計測対象の位相情報を持つことが多い。
その他、これらの重み付け処理は、ブラインド・デコンボリューションにおいて実施されることもある。また、別に求められた点拡がり関数やシステム伝達関数を用いて逆フィルタリングして白色化する場合や、それら点拡がり関数やシステム伝達関数の共役や、式（ＡＡ１）の共役を掛ける場合を含め、それらをビームフォーミング前又は後（送信ビームフォーミングのみの状態や開口面合成用に収得した受信信号）において実施する場合においても、これらの重み付け処理は有用である。特に、逆フィルタリングにおいては、正則化が施されることもある。また、最尤推定(MAP有り又は無し)等に基づいて実施されることもある。 For higher resolution, super-resolution in the linear model described above is effective, and these super-resolutions are sometimes used for such multiple waves and beams. That is, individual waves or beams that are not superimposed may be super-resolved and superimposed, or super-resolved while being superimposed. In some cases, both are mixed and processed. In some cases, noise is reduced by overlapping (additional averaging) under the same parameters and processed. Significantly higher spatial resolution can be achieved for the original signal (including harmonics). Although various super-resolutions have been described, for example, as described in paragraph 0363, when inverse filtering is performed with the desired point spread function or signal distribution spectrum such as echo distribution as a target, paragraphs 0390 to 0415 As described with the displacement measurement, the Wiener filter can be applied and weighted for the imaging of the signal itself. With the underlying weights (the squared norm of the original signal spectrum removed), the inverse filter

where Hp(ωx, ωy, ωz) and G(ωx, ωy, ωz) are the spectra of the signal to be processed and the target signal, respectively.
weighted by the norm of

where PWpn(.omega.x, .omega.y, .omega.z) and PWps(.omega.x, .omega.y, .omega.z) are the power spectrums of noise and signal, respectively. q is any positive value.
should be processed using Also in other super-resolution in the linear model described above, it is possible to similarly apply the Wiener filter to suppress noise amplification. When applying the norm of (AA1) itself without using a Wiener filter, ε (< ₁ ) times the norm of the signal spectrum Hp (ωx, ωy, ωz) ₍ maximum norm, L2 norm, L1 norm, etc.) In some cases, only the spectrum of frequencies having the above spectrum is processed (spectra of other frequencies are made zero). In these cases, (AA1) itself may be used irregularly instead of the norm of (AA1), and the phase may also be matched. In that case, Gp(ωx, ωy, ωz) often has the phase information of the measurement target.
Alternatively, these weighting processes may be implemented in blind deconvolution. In addition, when whitening by inverse filtering using a separately obtained point spread function or system transfer function, or when multiplying the conjugate of the point spread function or system transfer function, or the conjugate of formula (AA1), These weighting processes are useful even when they are performed before or after beamforming (transmit beamforming only or received signals acquired for aperture plane synthesis). In particular, regularization may be applied in inverse filtering. It may also be performed based on maximum likelihood estimation (with or without MAP).

The above-described nonlinear processing may be applied to a signal super-resolved under such a linear model. Furthermore, high resolution and high contrast can be achieved. As a result of the super-resolution obtained under the linear model, the super-resolved original signal (including the case of harmonics, hereinafter the same), and the super-resolved signals are Objects to be processed include those in which a plurality of originals exist and are superimposed, those super-resolved by superimposing a plurality of originals, and those in which they are mixed and processed. Also, there may be a plurality of original signals, each of which is super-resolved under a linear model, subjected to nonlinear processing, and superimposed. They are sometimes mixed and processed.

Also, as described above, the non-linearly processed signal may be subjected to super-resolution performed with a linear model. High resolution is achieved, but contrast may be reduced. As a result of super-resolution obtained by non-linear processing, the original signal (including the case of harmonics) is super-resolved, and multiple super-resolved signals exist and are superimposed The objects to be processed are those that have been superimposed, those that have been super-resolved by superimposing multiple originals, and those that have been mixed and processed. Also, there may be a plurality of original signals, each of which is super-resolved by non-linear processing, super-resolved by a linear model, and superimposed. They are sometimes mixed and processed.

In the same position of interest, when these nonlinear processes are applied to a plurality of signals (signals representing beams or waves, not limited to fundamental waves and may be harmonics), the aperture directivity and If the intensity of the original signal itself before processing differs due to the effects of scattering and attenuation (including frequency-dependent cases) in the target, the difference may be emphasized, especially higher harmonics. When generated, it can be noticeable. This difference may be positively imaged or quantitatively confirmed in a spectral image (their frequency characteristics may be emphasized and confirmed). On the other hand, in order to image with the difference reduced, before or after the nonlinear processing, processing for weighting the energy of the signal or the spectrum of a specific frequency may be performed for imaging. A plurality of these signals may be superimposed and imaged. It should be noted that the spectrum or energy of the signal may be evaluated in a local region containing the location of interest, or may be evaluated over the entire region of interest. Of course, regardless of whether or not non-linear processing is applied, weighting may be performed in the same manner during linear superimposition processing.
In addition, although the signal strength becomes weaker in the direction of distance due to the effects of attenuation and reflection/scattering during the propagation process, for example, plane waves are attenuated less than during focusing. When non-linear processing is applied, the higher the order, the more the effect is emphasized. Therefore, as described above, signal strength may be corrected before or after non-linear processing (and may also be corrected without non-linear processing). It is processed before or after detection.
There are various other methods of super-resolution, and at least one of them may be used in combination, applied to the same or different signals, coherently added, and used. As noted above, it may also be applied to the superimposed signals. They may also be added incoherently, reducing speckle. Incoherent addition through super-resolution may not reduce spatial resolution, as described above.
Moreover, in each of these super-resolution processes, addition may be performed spatially non-uniformly depending on the signal intensity, SN ratio, and spatial resolution. That is, the parameters of each method may be variable depending on those of each location. In the case of spectrum processing, it is as described above. For example, in the case of non-linear processing, it is the order of exponentiation, the number of multiplications, and the like. In addition, when performing addition of coherent signals and incoherent signals, it is the number of additions, weighting values, and the like. Of course, it may be processed spatially uniformly.
As an effect of increasing the contrast by nonlinear processing, scatterers and reflectors in the tissue can be visualized conspicuously. In some cases, the effect of making the difference in height (in brightness) noticeable may be obtained. For example, it may become easier to capture post-necrotic calcification of biological tissue. In addition, for example, they may be colored depending on their signal strength and displayed superimposed on a normal gray image, Doppler image, power Doppler image, contrast image, or the like. Processing may be performed after the intensity of the signal distribution is corrected. For example, the intensity of the signal received from within the region of interest (before or after detection) is corrected to be spatially uniform, and then subjected to nonlinear processing to obtain the scattering intensity distribution, the scattering intensity of a plurality of scatterers, or the reflection intensity distribution. In some cases, the high and low reflection intensities of multiple reflectors are visualized. Processing may be applied for the purpose of counting the number of reflectors or scatterers. In addition to focused beams and aperture plane synthesis, when plane waves, spherical waves, or cylindrical waves are used, the spatial resolution is low. In particular, when nonlinear processing is performed, scattered waves and reflected waves can be visualized remarkably at their generation positions. For example, when these crossing waves are generated, a cross-shaped waveform is emphasized and displayed at the scatterer position as the scattered wave.

We also calculated the exponentiation and multiplication of the point spread function when using a concave aperture HIFU applicator (simulation, frequency 5 MHz, aperture diameter 12 mm, focal depth 30 mm) (single aperture and dual aperture). . As noted above, this type of calculation is effective in considering heating effects. By collecting experimental data, it is possible to formulate the relationship between sound pressure (point spread function), harmonic sound pressure, heat reception, etc., and design applicators and radiation sound pressure (ultrasonic parameters). Through this, it can be used to improve the efficiency of heat treatment. The same is true when other waves are used.

FIG. 51 illustrates the sound pressure change according to one embodiment of the invention when using a concave HIFU applicator. In FIG. 51, (a1) and (a2) are images of the sound pressure due to the original signal and the sound pressure of the squared signal (left contains DC, right only harmonics) when using one aperture. shows an image of (b1) and (b2) are images of the sound pressure of the original signal and the sound pressure of the multiplied signal (left: DC , and the image on the right shows only harmonics), respectively. The image is by envelope detection and the image size is 3.8×12.8 mm ² . In the second harmonic components obtained by squaring and multiplication, it can be confirmed that the sound pressure is concentrated in the desired region and the contrast is high. In this way, it is possible to evaluate the intensity (intensity, that is, power) of the fundamental wave and estimate the sound pressure distribution shape of the generated harmonics (second or higher harmonics). Also, the power (Intensity) consumed by harmonics can be evaluated. The harmonics actually observed can be similarly evaluated.

In the above, regarding the imaging apparatus according to the embodiments of the present invention, mainly nonlinear operation devices for electromagnetic waves, vibrations including sound, heat waves, or corresponding signals have been described. enhancing effects, simulating non-linear effects, or virtually realizing non-linear effects (that is, non-linear effects other than those that physically, chemically, or biologically produce them (including the case where it does not occur) is also possible. In that case, the wave-related devices used can be used simultaneously to receive waves, or the invention can be implemented on the basis of signals received at different times if they are in phase. That is, the present invention can handle cases where multiple waves occur simultaneously and cases where waves occur singly.

In addition, in electromagnetic waves, vibrations, or heat, if the frequency is different, the dominant behavior will be different depending on each measurement object (medium), and it is true that the names are different (for example, for electromagnetic waves , microwaves, terahertz waves, radiation such as X-rays, and other vibrations. For example, in the case of human soft tissue, shear waves do not propagate as waves due to the effects of attenuation in the megaHz band, and ultrasonic waves are dominant. However, at low frequencies such as 100 Hz, the incompressible characteristics are strong and the shear wave is dominant), the present invention enhances the nonlinear effect between those with such different behaviors, and the nonlinear effect can be simulated or virtually realized.

In that case, the wave-related devices used can be used simultaneously to receive waves, or the invention can be implemented on the basis of signals received at different times if they are in phase. Of course, the propagation velocity of each wave, attenuation, scattering, reflection, and other phenomena have dispersion characteristics, and there is a limit that must be used appropriately in consideration of the SN ratio of the received signal. However, the range of applications of the present invention is very broad, including the ability to generate high frequency and low frequency signals that cannot be physically generated or captured.

In addition, it is described that the nonlinear operation and the nonlinear effect in the measurement object are imaged, or that it is applied to other measurements, and that harmonics are actively propagated to the measurement object. In some cases, the fundamental wave is actively used together in them. It is also possible to construct over-determined systems. Fundamentals are treated similarly to harmonics.

Furthermore, at least one of a plurality of signals obtained through nonlinear arithmetic is subjected to arbitrary detection processing, or multiple signals that may include a fundamental wave are subjected to arbitrary detection processing and then superimposed, or Arbitrary detection processing may be performed on a superimposition of a plurality of signals, which may include a fundamental wave, and imaging or other measurements such as displacement may be performed. Regarding superposition, the former incoherent addition (incoherent compounding) is effective in reducing speckle, and spatial resolution does not decrease when a high-frequency signal is generated and used. Low spatial resolution, which often occurs in normal speckle reduction, does not pose a problem. If a low frequency signal is generated and used, the spatial resolution is reduced but may be useful. On the other hand, the latter coherent addition (coherent compounding) can widen the bandwidth of the signal, that is, increase the spatial resolution. In particular, when a high frequency signal is generated and used, the frequency can be increased, and when a low frequency signal is generated and used, the frequency can be decreased. As a result, imaging can be performed with high spatial resolution, and displacement and other measurements can be performed with high accuracy. As described above, signals obtained by generating a plurality of beams or waves or by frequency division of a spectrum are also included in these processing targets including nonlinear processing.

Displacement measurements can be applied, for example, as described above. The range of applications is immeasurable, such as radar, sonar, and environmental measurement. In addition to displacement, for example, temperature may also be measured. Temperature sensors may be used directly for temperature sensing, or the temperature dependence of wave propagation characteristics may be detected. Temperature distribution may also be measured based on signal processing for the purpose of strain measurement. Chemical shifts in nuclear magnetic resonance frequencies may also be detected based on signal processing. It is also possible to observe heat waves, image their nonlinearity, and apply them to improve the efficiency of heat treatment.

It should be noted that the case of positively observing the nonlinear effect in the object to be measured and the case of positively applying the nonlinear processing according to the present invention may be used by switching between the two, using both at the same time, or actively using both. Non-linear effects within the measurement object may also be accounted for through simple computations. That is, by making full use of the nonlinearity of the signal source, the contrast agent, and the analog or digital nonlinear calculations used, the nonlinear effect within the object to be measured can be measured with high accuracy and imaged.

The above imaging and measurement are based on appropriate beamforming, and appropriate detection methods and tissue displacement measurement methods are also important. In the past, the inventors of the present application have used quadrature detection, envelope detection, square-law detection, and the like as detection methods for multidimensional received signals, and transverse modulation methods using cross beams as beamforming methods (Non-Patent Document 13 and 30), spectral frequency division methods (see Non-Patent Document 30), wave or beam shape tuning by spectral filtering, methods using many intersecting beams, and over-determined systems. Also, as a displacement vector measurement method, we have developed a multidimensional autocorrelation method, a multidimensional Doppler method, a multidimensional cross spectrum phase gradient method, and a phase matching method (see Non-Patent Documents 13 and 30). In addition, reconstruction imaging of (visco)shear elastic modulus distribution and thermophysical property distribution can also be performed based on displacement and strain measurements. Not only the original waves and signals, but also the superposition of multiple waves and signals, or the waves and beams (including pseudo ones) generated by nonlinear effects, pseudo waves in spectral frequency division and beams can be generated, and spectral filtering can adjust the shape of waves and beams.

A signal obtained by superimposing multiple signals that may contain a fundamental wave, or a signal obtained by spectrally dividing or filtering at least one of a plurality of signals that may contain a fundamental wave in the frequency domain ( See Non-Patent Document 30), or the original signal that has not been subjected to these processes, or a combination of these to configure an over-determined system, and the above processing, such as imaging , or other measurements such as displacement may be taken.

As described above, according to the present invention, a coherent signal obtained by detecting a transmitted wave, a reflected wave, or a scattered wave of an arbitrary wave by a sensor is superimposed with a nonlinear response to high intensity during wave propagation and a wave. Imaging using the original signal by obtaining non-linear effects such as multiplication and exponentiation (generation of harmonics, chords, difference tones, etc.), for example, by performing analog calculations or digital calculations using a computer It realizes high-frequency, wide-band, high-contrast, and high-spatial-resolution imaging. It is also possible to perform low-frequency imaging instead of high-frequency imaging. Also, under the same effect, measurement of displacement, velocity, acceleration, strain, or distortion rate can be realized with higher spatial resolution and higher accuracy than Doppler measurement using the original signal.

The superposition of the waves means what is physically achieved during beamforming, between beamformed waves, non-beamformed waves, and the like. When the wave intensity is weak, the superposition theory, which is established by the linear law, is mainly observed. waves, chords, and difference tones), and the present invention focuses on this latter phenomenon. The present invention is also characterized by the ability to use all of these wave components and superimposed waves regardless of their strength. Of course, it also applies to waves containing the fundamental wave and harmonic components artificially radiated or generated during propagation. Examples of harmonics generated during propagation include, for example, ultrasonic harmonic signals.

On the other hand, the present invention provides, for example, beams generated by beamforming (apodization, delay processing, or addition processing, either physical or computational), or waves that have not undergone beamforming. Regarding arbitrary waves such as plane waves and received signal groups for aperture plane synthesis, transmitted waves, reflected waves, and scattered waves, nonlinear effects due to high intensity and multiple waves propagating in the same direction or different directions Highly accurate measurement of nonlinear effects such as multiplication and exponentiation that occur when waves (same waves with the same physical quantity but different directions, waves with different parameters with the same physical quantity, waves with different types of physical quantities) overlap, etc. Alternatively, in order to simulate, for example, after the signal is detected by a transducer, the analog calculator or digital calculator is actively used to perform those calculations on the signal, and broadband harmonics, chords, and differences you can get the sound. In addition, physical, chemical, or biological non-linear effects may be emphasized, or non-linear effects may be produced when the non-linear effects cannot be observed or do not occur. It is possible. Also, a plurality of detection signals can be obtained simultaneously. In addition, the application of basebanded signals generated by physical action is also included in the present invention. or apply the signal estimated by calculation).

In the past, the inventor of the present application has disclosed a lateral modulation method using cross waves (such as plane waves) and cross beams based on linear laws (having carrier frequencies in the depth direction and the lateral direction). According to the invention, it is also possible to obtain exponentiation effects in this transverse modulation as well as multiplicative effects between crossing waves. Also, normally, power-law effects and product effects can be obtained by increasing the wave intensity, but according to the present invention, these nonlinear effects can be obtained regardless of the wave intensity.

In addition, according to the present invention, a basebanded signal can be obtained by a new detection process that can be easily performed with a small amount of calculation instead of normal quadrature detection or envelope detection, and the effect is improved in echo imaging and Doppler measurement. is obtained. However, since the detection signal contains direct current unlike the so-called baseband signal, it is used as it is or after removing the direct current by analog processing or digital processing. In addition, the present invention also includes applications of basebanded signals obtained by physical action. A baseband signal generated by calculation is also referred to as a basebanded signal.

For example, in the fields of medical ultrasound and sonar, nonlinear phenomena in the propagation process of ultrasonic waves in the living body It has been clinically applied under the name of harmonic echo imaging caused by the accumulated in the elec- tric field, but the application of physically generated basebanded signals is not disclosed. The application of basebanded signals that are physically generated by other nonlinear phenomena is also not disclosed. Baseband signals, basebanded signals, and incoherent signals subjected to envelope detection or square-law detection are also included in the processing objects of the present invention.

In particular, with respect to Doppler measurement, the inventors of the present application use a multidimensional received signal to measure displacement in the direction of wave propagation, unlike normal Doppler measurement. It has made it possible to measure the strain tensor or the strain rate tensor with high accuracy. According to the present invention, unlike normal detection, a signal (basebanded signal) detected in an arbitrary direction from a multidimensional received signal can be obtained simultaneously, and using a normal one-directional displacement measurement method, It is also possible to easily perform these measurements in a short time with a small amount of calculation. Also in this case, echo imaging using the simultaneously obtained harmonics and the base banded signal is possible. In addition, it is possible to suppress side lobes and increase contrast. The temperature measurement is also performed as described above.

The basic idea is that multiplication between different single-frequency sine and cosine signals produces chords and difference tones, and exponentiation raises the frequency of a signal by a power factor. Also, a signal (distorted wave) having a plurality of frequency components is not only increased in frequency but also broadened in bandwidth. In addition to this, an effect of suppressing so-called side lobes is obtained, and the contrast is increased. These effects are often observed as effects during propagation of particularly strong waves, but in the present invention, any signal is subjected to analog or digital arithmetic processing to enhance the nonlinear effects, regardless of the strength. or imitate it, or newly generate it. It can also be implemented virtually. Harmonics and detection signals can be obtained physically or artificially, not only in the case of spatial resolution, but also in continuous waves. If a physically generated basebanded signal can be understood under the present invention, its application will also be useful in engineering. For example, it may be possible to measure displacement or displacement vector components (normal one-dimensional displacement measurement methods can be used). In addition, the observed harmonics can be used to measure displacement and displacement vector components (various multidimensional displacement vector measurement methods described above can be used), and constitute over-determined systems. Similarly to the case of using the nonlinear processing of the present invention, it can be applied to imaging (higher SN ratio, higher resolution, speckle reduction, etc.), displacement component measurement (higher accuracy), and the like. In these cases, it is also effective to positively apply a contrast agent to enhance the nonlinear effect.

Other applications of the present invention include heating, heating, cooling, freezing, welding, repair, and heating of cancerous lesions in medical treatment using waves (laser, ultrasonic waves, or focused high-intensity ultrasonic waves, etc.). , cryotherapy, or cleaning of arbitrary objects (glasses, etc.), enhancing their effects through nonlinear phenomena, making them high resolution, and predicting their effects (e.g., powerful ultrasonic waves The effect is improved through not only the power law effect during heating using cross beams, high spatial resolution by addition when enhancing the effect using cross beams, but also high frequency and high spatial resolution as multiplication effect). becomes possible. Also in these cases, continuous waves are sometimes used, and the same effect can be obtained.

The present invention can also obtain a nonlinear effect under physical conditions where the nonlinear effect cannot be obtained physically (for example, when the strength of the measurement target cannot be increased or when a high strength cannot be obtained due to high frequency). Effective, but conversely, the present invention is practiced under conditions in which contrast agents such as microbubbles are used during ultrasound echo imaging, displacement measurement, temperature measurement, or therapy to enhance nonlinear effects. is also valid. That is, the present invention can enhance nonlinear effects, but can also be simulated, newly generated, or virtually realized. In addition, the present invention can also be implemented purely for the purpose of improving resolution, precision, and efficiency in imaging, displacement measurement, treatment, and the like.

According to the present invention, similarly to harmonic imaging, it is possible to increase the frequency, broaden the band, increase the contrast, or suppress sidelobes, thereby enabling nonlinear imaging with a high SN ratio. Additionally, analog detection or digital detection can be performed simultaneously without requiring much memory or computation.

The effectiveness of the present invention has been demonstrated through simulations and agar phantom experiments by performing ultrasonic echo imaging and measurement imaging. The present invention can be applied to any coherent signal other than the ultrasonic echo method (familiar ones include light waves, OCT, electricity, magnetic field signals, radiation such as X-rays, heat waves, etc.), and between different types of coherent signals can also be applied. Signals, including those made incoherent by analog processing (eg, energy detection of received signals using sensors, energy detection using non-linear elements, etc.) or digital processing, may be processed along with coherent signals.

On the other hand, in the field of image measurement, motion observation is often performed using a coherent signal converted to an incoherent signal (result display is an image) through various detections (physical phenomena or normal signal processing). known (cross-correlation processing, optical flow, etc.). If this method is used for incoherent signals, it is also possible to widen the bandwidth (improve the resolution). The above high-resolution detection signal obtained by the present invention can also be used. The measurement accuracy of those movements is also improved. That is, the invention applies to any coherent or incoherent signal.

Imaging and motion measurement using the above-mentioned coherent and incoherent signals have been performed in various fields including the above examples, and have a very long history. Alternatively, it is effective and useful to perform wideband, high-resolution and high-contrast imaging including low frequencies, and to measure displacement or the like with high accuracy. It is effective, including the engineering effect unique to multidimensional signal processing. In addition, in other applications such as medical treatment, it is also effective and useful to evaluate and apply the effects based on the imaging of the nonlinear effects.

In an imaging device, it is useful to perform multiplication or power calculation to generate harmonic components and a basebanded signal (the new detection signal described above), and to generate an image signal based thereon. The same effect can be obtained not only when high-order nonlinear processing is performed. Depending on the balance of cost and the like, it may be selectively adopted or used in combination with conventional technology.

The present invention is not limited to the above embodiments, and many modifications are possible within the technical concept of the present invention by those skilled in the art. Targeting electromagnetic waves, vibrational waves (mechanical waves) including sound waves (compression waves), shear waves, surface waves, etc., and waves such as heat waves, reflected waves, transmitted waves, scattered waves (forward scattered waves or backward scattered waves, etc.), refracted waves, surface waves, shock waves, those waves generated from self-emanating wave sources, waves emitted from moving bodies, or waves arriving from unknown wave sources, etc. Appropriately configured digital signal processing algorithms (in digital circuitry or software implementations), such as displacement (vector) measurements, temperature measurements, or analog or digital hardware are used for the observed waves. Patent Documents 5 to 7 and 11 regarding displacement measurement methods (obtaining the propagation direction of the sensing wave with high precision and observing the displacement of the object, the particle displacement, the medium displacement, etc. with high precision, and the propagation of the wave is the object of observation (Detailed high-precision observation of cases, etc.) and other measurement methods are often disclosed centering on examples of the reflection method and the echo method. can be used for, and is not limited to.

INDUSTRIAL APPLICABILITY The present invention can be used in a beamforming method for performing beamforming using arbitrary waves arriving from an object to be measured, and in a measurement imaging device and a communication device using such a beamforming method. be.

Wave signals such as radar, sonar, and other optical waves, sound waves, and heat waves are now often generated by digital devices. It can be applied in conjunction with signal generation only by providing the processing ability to perform the following calculation. Multidimensionalization of various devices is also being attempted, and the importance of the present invention will increase. The objects to be measured are solids, fluids, rheology, inorganic substances, organic substances, living things, environments, etc., and the range of measurement is expected to expand immeasurably. In the future, the size of each device in the apparatus will be further reduced, and a computer can be incorporated at a lower cost even if the computational stress is sufficiently high. You can expect a lot. In addition, it is expected that the development of applications through measurement using waves will become more active, and the range of applications will be broadened. Digitization of various devices is expected to progress further in the future, and at that time, it is believed that the demand for high-precision and high-speed beamforming capable of generating high-precision signals in real time according to the present invention will increase. Best of all, in addition to being faster, there is no need to do the interpolation approximations that were required heretofore. However, in the present invention, further high-speed processing is emphasized, and processing involving interpolation approximation may be performed as necessary, although the accuracy is degraded. It is also an effective device for normal communication and sensor networks. The availability and marketability of digital beamforming according to the present invention based on digital signal processing are more than sufficient.

DESCRIPTION OF SYMBOLS 10... Transmission transducer (or applicator), 10a... Transmission aperture element, 20... Reception transducer (or reception sensor), 20a... Reception aperture element, 30... Apparatus main body, 31... Transmission unit, 31a... Transmitter, 32... Reception Unit, 32a... receiver, 32b... AD converter, 32c... memory (or storage device or storage medium), 33... digital signal processing unit, 34... control unit, 35... receiving unit (or receiving device), 35a... reception device, 35b...AD converter, 35c...memory (or storage device or storage medium), 35d...phasing adder, 35e...other data generator, 40...input device, 50...output device, 60...external storage device, Reference Signs List 1 Measurement object 1a Contrast agent 110 Transducer 111 Nonlinear device 112 Working device 120, 120a, 120b Imaging apparatus body 121 Transmitter 121a Transmission delay element 122 Receiver 122a... Reception delay element, 123... Filter/gain adjustment unit, 124... Non-linear element, 125... Filter/gain adjustment unit, 126... Detector, 127... AD converter, 128... Storage device, 129... Reception beamformer, 130 Calculation unit 131 Image signal generation unit 132 Measurement unit 133 Control unit 134 Display device 135 Analog display device 140 External storage device

Claims (46)

A measurement imaging apparatus for generating an image signal of a measurement target when arbitrary waves are transmitted from a wave source positioned in an arbitrary direction toward the measurement target in a rectangular coordinate system defined by a reception aperture element array,
a plurality of reception aperture elements for receiving waves arriving from the object to be measured and generating respective reception signals;
(a) a beam to one of the received signal , the quadrature detection signal of the received signal, the IQ signal (in-phase signal and quadrature phase signal) of the received signal, and the analysis signal of the received signal; A complex exponential function whose exponent part is the product of a selected single angular frequency at a sampling instant near a desired instant to which the forming process is to be applied and the time difference between said desired instant and said sampling instant . or (b) a selected single angular wavenumber at a spatial sampling location in the vicinity of the desired spatial location to which the beamforming process is applied, and the desired spatial location and the spatial obtaining the signal at the desired time point or spatial position by interpolation approximation by applying at least a phase rotation based on multiplication of a complex exponential function whose exponent part is the product of the distance between the target sampling position and the plurality of Phased addition processing is performed using a plurality of signals obtained for the reception aperture elements, and steering represented by zero or non-zero deflection angles or elevation and azimuth angles with respect to the front direction of the plurality of reception aperture elements. Signal processing for generating the image signal in a desired coordinate system having coordinate axes in at least two directions by applying a transmit or receive component beamforming process including steering an angular transmit or receive component. Department and
A measurement imaging device comprising: 2. The metrology imaging apparatus of claim 1, wherein the exponential portion of the complex exponential function includes phase aberration correction. The signal processing unit is
(i) In order to perform phasing addition processing on the received signal, the sampled signal itself of the received signal and the spatio-temporal interpolation approximation processing on the sampled signal, or a zero spectrum around the spectrum of the sampled signal At least one of the up-sampled ones obtained by performing interpolation approximation processing by broadening the band by adding and performing phasing addition processing,
(ii) In order to perform phasing addition processing on the received signal, the spectrum of the local signal represented by the frequency corresponding to the sampling direction of the received signal is subjected to (a) angular frequency and beamforming processing. A phase rotation based on multiplication of complex exponential functions having in the exponent part the product of the time difference between the desired instant to apply and the time difference between the nearby sampling instant , where the exponent part may contain the correction , or (b ) a complex exponent having in the exponent part the product of the angular wavenumber and the distance between the desired spatial position to which the beamforming process is to be applied and the neighboring spatial sampling position, where the exponent part may contain the correction performing phase rotation based on multiplication of a function, and at least adding a plurality of local signals phased in the sampling direction for the plurality of reception aperture elements;
(iii) performing wavenumber matching in a Fourier transform or a subsequent inverse Fourier transform to perform Fourier beamforming on said received signal, without interpolation approximation in any coordinate axis direction, or at least one coordinate axis; performing wavenumber matching including interpolation approximation in the direction;
3. The metrology imaging apparatus of claim 1 or 2, generating another image signal different from said image signal in said desired coordinate system based on at least one of: at least one transmit aperture element that transmits at least one beam or at least one wave forming a steering angle expressed in zero or non-zero declination or elevation and azimuth angles toward the measurement target;
At the same time, at different times in the same time phase, or at substantially the same time phase, the declination or elevation and azimuth angles are the same, or at least one of the declination, elevation and azimuth angles is different. In addition, a plurality of beams or a plurality of waves in which other wave parameters and beam forming parameters including parameters related to focus or phase aberration are the same or at least one of which is different are directed toward the measurement target. at least one transmitting aperture element for transmission;
or at least one of said plurality of receive aperture elements doubles as said at least one transmit aperture element and generates a transmit signal that drives said at least one transmit aperture element 4. The measurement imaging apparatus according to claim 1, further comprising a signal generator. When scanning is performed with an effective aperture based on aperture plane synthesis processing for transmission or reception using a transmission aperture element array including a plurality of transmission aperture elements, the signal generation unit generates a plurality of transmission apertures to increase transmission intensity. The elements are simultaneously driven as one transmission aperture element to transmit waves, or the signal processing unit combines a plurality of reception aperture elements as one reception aperture element to synthesize received signals in order to increase the reception intensity. The measurement imaging apparatus according to claim 4. 6. The method of any one of claims 1 to 5, wherein in a single transmission or reception, transmit focusing is performed at multiple different locations in the same or different directions, or steering the transmit beam or receive beam in multiple different directions. 11. The measurement imaging device according to the item. The signal processing unit performs Hilbert transform by performing one-dimensional Fourier transform processing, multi-dimensional Fourier transform processing, or differentiation processing on the received signal, or performs quadrature detection by performing nonlinear processing. The measurement imaging apparatus according to any one of claims 1 to 6, wherein the analytic signal is obtained by: Another image signal processed by performing at least one additional processing before, during, or after the signal processing unit generates the image signal, and scattering, attenuation, transmission, and reflection. , refraction, impedance, velocity of propagation, direction of propagation, steering angle, interference, phase aberration, frequency dispersion, or nonlinear characteristics, and/or measurement data about the wave, and measuring the phase aberration performs phase matching, which may be iterative and may involve optimization, based on cross-correlation-based processing, autocorrelation processing, block matching, or phase rotation processing in space-time or frequency space; and , using the received signal, the image signal, the processed another image signal, or the measurement data about the wave, electromagnetism, mechanics, thermology, biology, or chemical phenomena related to the measurement object A metrology imaging apparatus according to any one of claims 1 to 7, capable of producing metrology data comprising composition, structure, displacement, temperature, or reconstruction comprising physical or chemical quantities or physical properties. (i) longitudinal wave, transverse wave, surface wave, incident wave, reflected wave, scattered wave, refracted wave, diffracted wave, interference wave, guided wave, shock wave, or , a wave that may include polarization, (ii) a wave source, (iii) a medium, and (iv) a wave transmitted or received for sensing of said measurement object. A received signal of transmitted or received waves, the image signal, another image signal processed by at least one additional processing, measurement data about the waves, for detailed measurement or analysis to obtain new information. Or, processing or processing measurement data in electromagnetic, dynamic, thermodynamic, biological, or chemical phenomena related to the measurement object in the spatio-temporal domain, wavenumber domain, or frequency domain, and measuring or The measurement imaging apparatus according to any one of claims 1 to 8, which is used for imaging. whether the wave itself transmitted from the wave source toward the measurement object is frequency-modulated in at least one direction in the spatio-temporal domain or the frequency domain;
Alternatively, waves transmitted from the wave source toward the measurement object are frequency-modulated in at least one direction in the measurement object,
Alternatively, the signal processing unit may process the received signal, the image signal, another image signal processed by performing at least one additional processing, measurement data related to waves, or electromagnetism, mechanics, The measurement according to any one of claims 1 to 9, wherein measurement data in thermodynamic, biological, or chemical phenomena are subjected to frequency modulation processing in at least one direction in the spatio-temporal domain or the frequency domain. imaging device. The signal processing unit performs, in the frequency domain, the received signal, the image signal, another image signal processed by performing at least one additional processing, measurement data related to waves, or electromagnetics related to the measurement target, Determining the upper limit of the steering angle from the band of the angular spectrum of the measurement data in dynamics, thermodynamics, biology, or chemistry, and filtering out the aliased spectrum part when aliasing occurs at that time or perform band widening through spectral zero stuffing in the direction in which aliasing occurs, or perform oversampling or upsampling in the direction in which aliasing occurs in the spatial domain, according to any one of claims 1 to 10 The metrology imaging device described. the received signal, the image signal, another image signal processed by performing at least one additional processing, measurement data related to waves, or electromagnetics, mechanics, thermology, biology related to the measurement object, or When the measurement data in the chemical phenomenon is expressed in an orthogonal coordinate system including a Cartesian orthogonal coordinate system or a polar coordinate system, the signal processing unit performs cross-correlation-based processing, autocorrelation processing, block matching, or spatio-temporal or It may be iterative to perform phase rotation in frequency space or phase itself, including phase aberration correction, translation processing, phase matching, alignment, position correction, displacement measurement, or motion compensation. In terms of translation and rotation in space-time, whether only the translation or the rotation is applied, the rotation is applied after the translation, the translation is applied after the rotation, or each is applied repeatedly or alternately 12. The measurement imaging apparatus according to any one of claims 1 to 11, which intersects expansion and contraction processing. wherein the signal processor corrects a waveform for a received signal that includes waveform distortion that may have frequency dependence caused by phase changes in the wave source or devices in the metrology imaging apparatus;
Alternatively, the wave source generates a wave that compensates for waveform distortion that may have frequency dependence occurring in the wave source or device that has been determined in advance, or
Alternatively, the measurement imaging apparatus according to any one of claims 1 to 12, wherein the signal processing unit corrects waveform distortion or attenuation of signal intensity occurring in the measurement target in addition to those waveform distortions. The signal processing unit processes the received signal, the received signal during beamforming processing, the image signal, another image signal processed by performing at least one additional processing, measurement data related to waves, or the measurement target for measured data on electromagnetic, mechanical, thermal, biological, or chemical phenomena relating to
linear processing, which may include superposition or spectral segmentation, which may include machine learning;
non-linear processing, which may include multiplicative or exponentiation effects, which may include machine learning;
amplification or attenuation, which may involve machine learning;
apodization, which may include machine learning;
filtering, which may include machine learning;
encoding or decoding or matched filtering or correlation processing for chirp wave transmissions, which may include machine learning;
detection, which may include machine learning;
detection processing, which may include said decoding, said matched filtering, said correlation processing, Wiener filtering, MUSIC (Multiple Signal Classification: a wireless communication technique using eigenvalues or eigenvectors of a correlation matrix of received signals), or machine learning;
The superimposition, the decoding, the matched filtering, the correlation processing, MIMO (Multiple-input and Multiple-output: wireless communication technology that combines multiple antennas on the transmitting side and the receiving side to widen the band of transmitted and received signals), SIMO (Single-input and Multiple-output: wireless communication technology that uses one antenna on the transmitting side and multiple antennas on the receiving side to widen the band of the received signal), performing the MUSIC, regularization or singular value decomposition Multiplexed processing of common information or independent information or signal source separation processing,
Phase, including block matching, phase processing, or machine learning, which may be iterative in performing cross-correlation-based processing, autocorrelation processing, block matching, or rotation of phase in space-time or frequency space aberration correction, movement processing, phase matching, registration, position correction, displacement measurement, or motion compensation;
image processing, which may include enhancement, machine learning, or image measurement;
Using excess systems under multiple observations as well as single observations, machine learning, regularization, Bayesian estimation, MAP (Maximum a posteriori) estimation, maximum likelihood estimation, Wiener filtering, least squares method , EM (Expectation-Maximization), singular value decomposition, linear or non-linear programming, or convex projection;
machine learning on coherent or incoherent signals, inverse filtering, Inverse Synthetic Aperture Radar (ISAR), said optimization processing, matched filtering, correlation processing, or super-resolution, which may include MUSIC;
resolution enhancement or contrast enhancement, which may include using machine learning, non-linear processing, or coherence factors;
speckle reduction or noise reduction, which may include machine learning;
adaptive beamforming, minimum variance beamforming, Capon method, or compressed sensing, which may involve machine learning or may perform regularization or singular value decomposition;
analysis or analytics, which may include machine learning;
Measurement or reconstruction (inverse analysis or inverse problem) of physical quantities, chemical quantities, or physical properties in electromagnetic, mechanical, thermal, chemical, or biological phenomena, which may involve machine learning or said optimization process When,
visualization, which may include machine learning;
mathematics or information processing, which may include stochastic statistical processing or machine learning;
feature extraction, instrumented equalizer generation, sparse modeling, data compression, or data mining, which may include machine learning;
parallel processing, which may include machine learning;
Mixing multiple homogeneous or heterogeneous information, which may involve machine learning;
Integration and fusion of multiple homogenous or heterogeneous information, which may include KL (Kullback-Leibler) information, mutual information, entropy, maximum likelihood estimation, Bayesian estimation, or EM methods, or may involve machine learning , quantifying, recognizing, judging, or diagnosing,
and performs at least one additional processing of the following: is equivalent to the controller of the surrounding device, communicates with another device, controls another device, is controlled by another device, or is a robot 14. The measurement imaging apparatus according to any one of claims 1 to 13, which may be installed in a functional device which may include a movable body. The received signal, the image signal, the processed separate image signal, the measurement data on the wave motion, or the measurement data on electromagnetic, dynamic, thermodynamic, biological, or chemical phenomena on the measurement object, When the signal processing unit has a low signal-to-noise ratio, contains multiple reflections or multiple scattering signals, or contains speckle signals, the signal processing unit performs cross-correlation-based processing, autocorrelation processing, block matching, or spatio-temporal processing. Or iteratively when performing phase rotation processing in frequency space, phase aberration correction, translation processing, phase matching, alignment, position correction, displacement measurement, motion compensation, or inference for different waves. Signals subjected to said filtering, said decoding, said matched filtering, said correlation processing, said Wiener filtering, said MUSIC, said detection processing, said separation processing, or said machine learning, which may include superposition of coherent signals. 15. The metrology imaging apparatus of claim 14, which performs detection. The signal processing unit is
(i) the received signal, the signal during beamforming, the image signal, another image signal processed by at least one additional processing, measurement data related to waves, or electromagnetics related to the measurement object; For measurement data in dynamics, thermology, biology, or chemistry, find the frequency related to the spatial position of the one-dimensional signal or multi-dimensional signal, and display the frequency itself or phase rotation and displaying the phase itself, multiplying it by a cosine function or sine function, or further weighting it with the envelope of the one-dimensional signal or the multi-dimensional signal and displaying it When,
(ii) one-dimensional or multi-dimensional received signal, image signal, processed another image signal, measurement data on the wave, or electromagnetism, dynamics on the measurement object, for the entire region of interest or for each local, Inverse Fourier multiplication of the frequency response of the measured data in a thermodynamic, biological, or chemical phenomenon, at each frequency, by the conjugate of the one-dimensional or multi-dimensional frequency response of the overall or local wave or wave properties of interest converting and displaying;
(iii) one-dimensional or multi-dimensional received signal, image signal, processed another image signal, measurement data related to the wave, or electromagnetics, dynamics related to the measurement target, for the entire region of interest or for each local, The frequency response of the measured data in a thermal, biological, or chemical phenomenon is divided at each frequency by the one-dimensional or multi-dimensional frequency response of the wave or wave characteristics for the entire region of interest or for each locality. obtaining a frequency characteristic and displaying its inverse Fourier transform;
(iv) one-dimensional or multi-dimensional received signal, image signal, processed another image signal, measurement data related to the wave, or electromagnetics, dynamics related to the measurement target, for the entire region of interest or for each local, Deconvolution of measurement data itself in thermodynamic, biological or chemical phenomena with one-dimensional or multi-dimensional impulse responses of waves or wave characteristics for the entire region of interest or for each local area to determine the characteristics of the medium. displaying the results;
to determine the spatial distribution or spatio-temporal distribution of the characteristics of the waves coming from the object to be measured, or the waves transmitted or received for sensing the wave source or medium or the object to be measured. or display with super-resolution, wherein the frequency characteristics or impulse response of the wave is obtained from another reference, the received signal, the image signal, or the processed another image signal A metrology imaging apparatus according to any one of claims 1 to 15, which can be obtained. the received signal, the image signal, another image signal processed by performing at least one additional processing, measurement data related to waves, or electromagnetics, mechanics, thermology, biology related to the measurement object, or A desired point spread function or spectral intensity of measurement data in a chemical phenomenon is subjected to beamforming under ideal parameters for the measurement target or reference object, or beamforming is performed multiple times and ensemble averaged. or perform averaging under a local stationary process and observe, or set by analysis, simulation, or optimization, and perform convolution processing on the result of deconvolution, or spectral intensity 17. The measurement imaging apparatus according to any one of claims 1 to 16, wherein the weighting process is performed on . the received signal, the image signal, another image signal processed by performing at least one additional processing, measurement data related to waves, or electromagnetics, mechanics, thermology, biology related to the measurement object, or Estimate a one-dimensional point spread function by obtaining a one-dimensional autocorrelation function or estimate a multidimensional point spread function by obtaining a multidimensional autocorrelation function for measurement data in a chemical phenomenon, (i) the wavelength of the wave, (ii) the propagation velocity of the wave, (iii) the radiation pressure or shock wave that is the force source of the mechanical wave, and (iv) the high intensity focused ultrasound that is the heat source of the thermal wave. estimating at least one of Ultrasound (HIFU) treatment pressure and (v) the shape of a wave source, including a self-emanating wave source, and applying an inverse analysis to the wave source. estimating the intensity, or measuring data about the received signal, the image signal, the processed another image signal, the wave, or electromagnetics, mechanics, thermology, biology, or 18. The metrology imaging apparatus according to any one of claims 1 to 17, wherein the metrology imaging apparatus is capable of estimating the spatial resolution of metrology data in chemical phenomena. For Doppler observation of a displacement vector, the signal processing unit processes the received signals of a plurality of intersecting beams or waves, the image signal, another image signal processed by performing at least one additional processing, the wave Aliasing occurs when measurement data or measurement data in electromagnetic, dynamic, thermodynamic, biological, or chemical phenomena related to the measurement object is subjected to digital demodulation processing or nonlinear processing to cause aliasing. Doppler processing is performed based on the carrier frequency directly and correctly obtained by replacing the frequency domain of the frequency coordinates in the direction in which aliasing occurs without broadband processing the received signal in the direction. 1. The measurement imaging device according to claim 1. The signal processing unit processes the received signal, the image signal, another image signal processed by performing at least one additional processing, measurement data related to waves, or electromagnetics, mechanics, and thermodynamics related to the measurement target , Biology, or performing processing to remove the side lobes or grating lobes of the measurement data in the phenomenon of chemistry, or actively using the side lobes or the grating lobes, any one of claims 1 to 19 4. The measurement imaging device according to . The signal processing unit is
(i) from the received signal, the image signal, another image signal processed by performing at least one additional processing, or the wave measurement data itself, in order to improve the accuracy of the imaging of the measurement object; using directly estimated variances or covariances or spectra;
(ii) In order to improve the accuracy of the measurement or reconstruction of the measurement target, use the variance or covariance or spectrum directly estimated from the actually measured measurement or reconstruction result itself, Alternatively, variance or covariance including Ziv-Zakai Lower Bound (ZZLB) or power Doppler indirectly estimated from the received signal, the image signal, or the processed separate image signal or optimization or stochastic processing in temporal or spatial coordinates using the spectrum;
The measurement imaging apparatus according to any one of claims 1 to 20, which performs at least one of The signal processing unit performs unbiased regularization for measurement or reconstruction, or stabilizes by setting a very large regularization parameter when the object of measurement or reconstruction has a constant value. 22. The measurement imaging apparatus according to any one of claims 1 to 21, wherein the accuracy is improved by A metrology imaging apparatus according to any preceding claim, wherein the waves received comprise at least one of any electromagnetic wave, any mechanical wave, and any thermal wave. 24. The measurement imaging apparatus according to any one of claims 1 to 23, wherein the wave motion to be received is a type of wave motion physically different from the wave motion to be transmitted toward the object to be measured. The signal processing unit includes a physical wave source, a physical transmission aperture element array, a virtual wave source set separately from the physical wave source, and a set separate from the physical transmission aperture element array. Processing the received signal with respect to a virtual transmit aperture element array, a physical receive aperture element array, or a virtual receive aperture element array set separately from the physical receive aperture element array, according to claims 1- 25. The metrology imaging device according to any one of 24. The signal processing unit performs displacement observations or predicts displacements based on models, which may include Markov models, to improve imaging or measurement accuracy using multiple waves or measurement data, and block matching. Or, it may be performed iteratively when performing phase rotation processing in spatio-temporal or frequency space, and performing phase aberration correction, translation processing, phase matching, registration, position correction, displacement measurement, or motion compensation. Then, the plurality of waves or measurement data are aligned, and the received signals of the plurality of waves of the same or different types coming from the object to be measured, the image signal, and at least one additional process are processed. In another image signal, measurement data related to waves, or measurement data in electromagnetic, dynamic, thermodynamic, biological, or chemical phenomena related to the measurement object, after performing the alignment, the same or different multiplexing of common information and separation of independent information or independent signal sources, super-resolution or image processing, or registration after super-resolution or image processing The metrology imaging apparatus of any one of claims 1-25, wherein As a process of integrating or fusing multiple events with different stochastic processes, it is possible to mix them, increase the accuracy by multiplexing, separate them into independent information, or transition to different stochastic processes by machine learning (probabilistic model or a change including a random variable). In machine learning that analyzes or analyzes the result of the integration or the fusion, or uses multiple teaching data of different types and multiple models of the same or different types, multiple models, multiple 14, 15, analyzing or analyzing the weight data of a model in which models are combined and trained at the same time, or a combined model in which a plurality of models each trained to some extent are combined and further trained; 26 or 27, the measurement imaging device. In machine learning, optimization, or in silico processing, data is encoded and processed, or input data for measurement or recognition is transformed into other homogeneous data, heterogeneous data, related data, related images , related recognition objects, numerical data, results of recognition or judgment or diagnosis, desired purposes, goals, physical properties, or functions, or bi-directional mapping in them, or of materials, structures, or functions A metrology imaging apparatus according to any one of claims 1 to 28 for obtaining a generating process or for interfacing with surrounding devices. Surgery that minimizes invasiveness under the differentiation between diseased tissue and normal tissue, precise high intensity focused ultrasound (HIFU) treatment of diseased tissue, electromagnetic wave heat coagulation treatment, radiotherapy, heavy ion beam In order to carry out treatment, physical therapy, chemotherapy, drug delivery treatment, or pharmaceutical therapy, ultrasonic echo equipment, optical ultrasonic equipment, electroencephalograph, superconducting quantum interference device (Superconducting Quantum Interference Device: SQUID), MRI (Nuclear at least one of a magnetic resonance imaging) device, an X-ray device, a terahertz device, a PET (positron emission tomography) device, a nuclear medicine imaging device, and an OCT (optical coherence tomography) device, or them It is possible to diagnose diseases of single or multiple organs simultaneously multilaterally or under the same index using one device configured by integrating or fusing multiple devices in the device, or 30. The measurement imaging apparatus according to any one of claims 1 to 29, which can be used before, during, and after treatment to perform integrated diagnostic imaging. When the measurement object includes inorganic matter and organic matter, the inorganic matter is observed using terahertz waves or X-rays, and ultrasonic waves, optical ultrasonic waves, MRI (nuclear magnetic resonance imaging) waves, or OCT (optical coherence tomography) 31.) The measurement imaging apparatus according to any one of claims 1 to 30, which observes organic matter using waves and displays the observation results by integrating or fusing them. The signal processing unit includes ultrasonic waves and optical ultrasonic waves that are mechanical waves, and MRI (magnetic resonance imaging) waves that are electromagnetic waves, OCT (optical coherence tomography) waves, laser light, microwaves, terahertz waves, and a Doppler measurement using at least one of X-rays. By analog or digital linear processing or non-linear processing in the signal processing unit, or by multiple waves or different types of waves with single or different wave or beamforming parameters inside or outside the measurement object and linear A metrology imaging apparatus according to any one of the preceding claims, wherein the metrology imaging apparatus receives waves affected by phenomena or non-linear phenomena to produce signals representing waves having new waves or beamforming parameters. When at least one wave that may be focused is transmitted from the wave source or at least one transmit aperture element, the signal processing unit processes the received signal, the received signal during beamforming, the image signal, Generating another image signal processed by applying at least one additional processing, measurement data on waves, or measurement data on electromagnetic, mechanical, thermal, biological, or chemical phenomena related to the object to be measured. , may be iterative during cross-correlation-based processing, auto-correlation processing, block matching, or phase rotation processing in spatio-temporal or frequency space, and may be super-resolved and then superimposed (compounded). processing), or perform super-resolution processing after superimposition processing,
The super-resolution processing is linear or non-linear processing, or a combination of linear and non-linear processing, and the superimposition processing is performed on a coherent signal to have the effect of increasing the resolution, or after detection has the effect of speckle reduction being performed on an incoherent signal of , or said non-linear processing has the effect of exponentiation, the processing of addition and subtraction of the results received under pulse inversion transmission 34. The metrology imaging apparatus of any one of claims 1 to 33, generating even and odd harmonics of a fundamental wave. The signal processing unit observes the measured wave motion or the dynamics or state of the medium based on Doppler processing or image processing for the measured wave motion, and determines the wave propagation velocity. When observing the anisotropy,
(i) propagating the wave in at least one desired direction to transmit or receive the wave for sensing and Doppler observation;
(ii) By propagating waves in multiple different directions at different times, transmitting or receiving waves for sensing each wave, superimposing the results of Doppler observation, and artificially generating waves propagating in new directions. generating or transmitting or receiving and superimposing waves for sensing each wave and observing Doppler;
(iii) Doppler observation by propagating waves in a plurality of different directions at the same time, transmitting or receiving waves for sensing superposition of those waves, or at least one Doppler observation filtering, separating or dividing the spectrum represented in wavenumber space or frequency space with respect to one wave;
35. The metrology imaging apparatus of any one of claims 1-34, wherein observations are obtained based on at least one of: The plurality of reception aperture elements or the signal processing unit observes audible sound waves, ultrasonic waves, compression waves, shear waves, vibration waves, seismic waves, or heat waves using electromagnetic waves including radiation, light, or microwaves. 36. A metrology imaging apparatus according to any one of claims 1 to 35 which uses ultrasound to observe audible, compression or shear waves. at least one of microwave, terahertz wave, or infrared radiation, shear modulus, viscoelastic modulus, nuclear magnetic resonance, ultrasonic waves, optical ultrasonic waves, and temperature dependence of physical properties of light 37. The measurement imaging apparatus according to any one of claims 1 to 36, wherein imaging of the temperature distribution inside said measurement object is performed based on one. At least one high-intensity focus of finite length from a transmit aperture element array comprising a plurality of transmit aperture elements to the subject, marked with a contrast agent or marker, or mechanically diagnosed. transmit ultrasound (High Intensity Focus Ultrasound: HIFU) and receive transmitted or reflected ultrasound from the subject, or at least one finite length ultrasound for observation from the transmit aperture element array is transmitted to the treatment target, and a received signal is generated by receiving transmitted or reflected ultrasonic waves from the treatment target,
In the signal processing unit, the treatment target under displacement measurement that may be performed repeatedly when performing cross-correlation-based processing, autocorrelation processing, block matching, or phase rotation processing in spatio-temporal or frequency space while tracking
Perform imaging or measurement, or calculate the phase aberration that occurs when steering in a direction significantly different from the front direction of the transmission aperture due to the spatial inhomogeneity or temperature dependence of the speed of sound or the directivity of the transmission wave, Phase aberration correction of HIFU transmissions to improve treatment location accuracy or cross-correlation-based processing, autocorrelation processing, block matching, or rotation of phase in space-time or frequency space in imaging or metrology. 38. A metrology imaging device according to any one of claims 1 to 37, wherein the metrology imaging device is fused with a HIFU treatment device, applying phase aberration correction, which may be iteratively performed in performing the to improve accuracy. while the signal processor tracks under displacement measurements a treatment object marked with a contrast agent or marker or mechanically diagnosed;
Thermal conductivity, heat capacity, or thermophysical properties including perfusion and heat source measurement data, or high intensity focused ultrasound (High Intensity Focus Ultrasound: HIFU) heat source database associated with beam forming parameters is subjected to a temperature distribution calculation simulator to predict the temperature distribution, or along with the monitoring of the temperature measurement, dynamic measurement of the treatment target and its surroundings Based on measurement imaging using mechanical properties, contrast agents, markers, therapeutic agents, or contrast agents or markers and therapeutic agents including shear modulus or viscoelastic modulus obtained under the monitoring of By performing non-invasive monitoring diagnostics on effects (degeneration including coagulation) and their surrounding tissue,
Performing an iterative heating treatment plan, with or without phase aberration correction, HIFU focus (irradiation) position, apodization (irradiation shape), irradiation intensity, continuous irradiation time , the irradiation interval, and the implementation interval are electronically or mechanically controlled in a switchable automatic mode or manual mode, and fused with the HIFU therapy device. 3. The measurement imaging device according to . The signal processing unit compares the results of the treatment performed under the prediction result and the prediction result as the sequential update type heating treatment plan, High Intensity Focus Ultrasound (HIFU) or another The treatment parameters are determined by linear or non-linear optimization or planning, wherein the comparative index is tissue physical properties including tissue pressure, temperature, exposure dose, shear modulus, or viscous modulus, 40. The metrology imaging device of claim 39, fused with a therapy device that references representative data, actual measurements, or models of heat-receiving or degenerative properties of a treatment subject. Detecting, diagnosing, or treating lesions by irradiating ultrasonic waves, other mechanical waves, or electromagnetic waves to a diagnostic site in order to use a magnetic substance that has affinity for human tissue as a contrast agent. 41. The measurement imaging device according to any one of 1 to 40. In order to separately observe different compositions or structures or tissues, different behaviors, at least one contrast agent, marker, or therapeutic agent present in the measurement object, in photoacoustic imaging, light absorption frequency Deterministically or stochastically observe the difference in the signal intensity or frequency characteristics of photoacoustic waves generated with respect to the characteristics or irradiated light, and the difference in signal intensity or frequency characteristics in other wave imaging. Based on the process of separating by using, or image processing or separation processing including deep learning,
(i) separating said received signal, said image signal, or another image signal itself processed with at least one additional processing, such that said different composition or structure or texture, said different behavior; , identifying a region, location, or boundary of the at least one contrast agent, marker, or therapeutic agent;
(ii) isolating measured displacement, temperature, or other measurement data from the received signal, the image signal, or the processed further image signal;
(iii) separating the received signal, the image signal, or the processed separate image signal and measuring displacement, temperature, or other measurement data;
42. The metrology imaging apparatus of any one of claims 1-41, performing at least one of: In photoacoustic imaging, a plurality of light sources or ultrasonic sensors having the same broadband characteristics or different band or frequency characteristics are arranged in a one-dimensional, two-dimensional, or three-dimensional array, ,
The plurality of light sources or ultrasonic sensors are locally continuously, alternately, or periodically arranged; Another image signal processed by applying two additional processes, filtering to select a frequency band or frequency for broadband observation, and superposition processing for observation of a specific band or frequency 43. The measurement imaging apparatus according to any one of claims 1 to 42, which selects an object to be imaged or measured by broadening the band. The measurement according to any one of claims 1 to 43, wherein at least one measurement target is observed from multiple angles in a non-contact in situ state, or simultaneously observed from multiple angles. imaging device. Multiple different data at the same event or at the same time, multiple independent data, data of multiple different frequency components or multiple different band components, data in multiple different frequencies or bands, or multiple different measurement targets or markers and display each image, numerical value, or processed result at the same time; 45. The measurement imaging apparatus according to any one of claims 1 to 44, wherein the image is displayed by superimposing with density or the image is displayed with watermark. Generate data on function, physical properties, structure, or composition for security, environmental observation, earth observation, meteorological observation, resource exploration, seismic observation, astronomical/space observation, underwater observation, or observation of structures , remote sensing, non-destructive testing, diagnosis, inspection, treatment, repair, creation, manufacturing, monitoring, monitoring, or biometric authentication including fingerprint identification or fingerprint authentication or iris authentication. The measurement imaging device according to any one of claims 1 to 3.

Images