JP2011030673A - Magnetic resonance imaging apparatus and image reconstruction program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a magnetic resonance imaging apparatus performing a highly precise estimation process suited to a radial sampling method. <P>SOLUTION: This magnetic resonance imaging apparatus acquires, in a measurement space, a plurality of nuclear magnetic resonance signal data which are radially disposed around an origin. In the measurement space, a filter is set for selecting data in a first region within a prescribed range from the origin and in a second region including a range separated from the origin with respect to the first region. The selected data are inverse Fourier transformed to acquire first-region inverse-Fourier-transformed data and second-region inverse-Fourier-transformed data. A phase value of the first-region inverse-Fourier-transformed data is subtracted from that of the second-region inverse-Fourier-transformed data to acquire complex conjugate data and acquire estimation data. A reconstruction image is acquired by adding the first and second inverse-Fourier-transformed data and the complex conjugate data. The first and second regions are set not to be overlapped with the estimation data, when the estimation data are Fourier transformed and virtually disposed in the k-space. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention measures magnetic resonance imaging (hereinafter referred to as NMR signal) from hydrogen, phosphorus, etc. in a subject and images nuclear density distribution and relaxation time distribution (hereinafter referred to as MRI). In particular, the present invention relates to an NMR signal estimation technique using a radial sampling method.

  An MRI apparatus measures an NMR signal generated by nuclear spins constituting a subject, particularly a human tissue, and images its head, abdomen, limbs, etc. in two or three dimensions. Device. In imaging, the echo signal is given different phase encoding depending on the gradient magnetic field and is frequency-encoded and measured as time-series data. The measured NMR signal is arranged in a measurement space (generally called “k space”, hereinafter referred to as k space), and is reconstructed into an image by two-dimensional or three-dimensional Fourier transform. . An NMR signal is acquired as an FID signal or an echo signal, but since it is almost always acquired as an echo signal, the NMR signal is hereinafter also referred to as an echo signal.

  The Cartesian sampling method is the most common method for arranging echo signals in the k-space. The Cartesian sampling method is a method in which a plurality of echo signals having different phase encodings are arranged in a direction parallel to the frequency encoding.

  There is a radial sampling method as one of methods for arranging echo signals in the k-space other than the Cartesian sampling method. The radial sampling method is a method in which a plurality of echo signals having different rotation angles with respect to the origin of the k space are arranged in a radial pattern. A pulse sequence of the radial sampling method is described in Patent Document 1, for example. In the radial sampling method, the echo signal for each angle is called Blade data.

  In the radial sampling method, (1) one echo signal (Blade) with no phase encoding added is arranged for each rotation angle, and (2) an echo signal group (Blade) with phase encoding added for each rotation angle. There are two types of arrangement (for example, Non-Patent Document 1). In the present specification, when distinguishing between the two, (1) is called a radial scan, and (2) is called a hybrid radial scan. When both are described comprehensively, it is called a radial sampling method. In the radial sampling method, since echo signals are arranged radially, it is said that the body motion artifact of the subject at the time of measuring the echo signal is less noticeable than the Cartesian sampling method.

  On the other hand, depending on the purpose of measurement, there is a sampling method in which echo signals are arranged only in a limited area in the k space. The purpose of arranging echo signals only in a limited area includes (1) shortening of TE (Echo Time) and (2) shortening of measuring time. In the case of (1), a fractional echo (Fractional Echo) signal from which the previous data on the time series of the echo signal is missing is measured by the MRI apparatus (hereinafter, such measurement is referred to as “fractional echo measurement”). In the case of (2), the number of echo signal measurements is reduced, and data in a specific area is lost depending on the pulse sequence (hereinafter, such measurement is referred to as fractional scan measurement). In a state where a part of such k-space data is missing, truncation artifacts and aliasing are superimposed on the reconstructed image. In order to suppress these artifacts, it is necessary to estimate and form missing area data (hereinafter, “estimated data” refers to data that estimates missing areas in k-space, and “estimates the process of calculating estimated data”. Process ”). As an existing estimation process, methods as shown in Non-Patent Document 2 and Patent Documents 2 and 3 below have been proposed.

特許文献２に記載の手法では、ハイブリッドラディアルスキャンにおける各々のBladeデータ内で位相エンコード方向にホモダイン処理が行われている。ホモダイン処理は、上記した非特許文献２の推定処理と同類の処理であるが、位相誤差の補正が加えられている。当該手法により、ＴＥ及び計測時間の短縮が可能となる。但し、位相補正の詳細に関しては特許文献２の中では詳細は記されていない。 Non-Patent Document 2 proposes an estimation method using the complex conjugate property. In this method, estimation processing is performed using the fact that the data in the k space is in a complex conjugate relationship. When the reconstructed image data is a real number, the data on the k space has a complex conjugate relationship as shown in Expression (1). Due to this property, echo signal data F (kx, ky) is measured only in half of the area in k-space (two quadrants not on the diagonal), and non-measured echo signal data F (-kx, -ky) ) Can be estimated from equation (1). However, in the formula (1), CC () indicates a complex conjugate in parentheses ().

In the method described in Patent Document 2, homodyne processing is performed in the phase encoding direction in each Blade data in the hybrid radial scan. The homodyne process is a process similar to the estimation process of Non-Patent Document 2 described above, but with phase error correction added. By this method, TE and measurement time can be shortened. However, details of phase correction are not described in Patent Document 2.

  The methods of Non-Patent Document 2 and Patent Document 2 described above are processing on a two-dimensional k-space, that is, the kx-ky plane. A process is disclosed. By performing estimation processing in a three-dimensional space, the number of echo signal measurements can be reduced as compared with estimation processing on a two-dimensional plane. Patent Document 3 discloses that a phase map is created from a part of k-space data and conjugate symmetric data is corrected.

International Publication WO2005 / 023108 JP 2006-223864 A Japanese Patent No. 3872016

Pipe JG et al., Multishot Diffusion-Weighted FSE Using PROPELLER MRI, Magnetic resonance in medicine, Vol. 47, No. 1, pp. 42-52, (2002) David A. Feinberg et al., Halving MR Imaging Time by Conjugation: Demonstration at 3.5kG, Radiology, Vol161, No2, pp527-531, (1986)

However, each of the estimation processes described above has the following problems.
The estimation method of Non-Patent Document 2 using the property of complex conjugate is based on the premise that the reconstructed image data is a real number, but the measured data has a phase error due to factors such as magnetic field distortion and noise. The reconstructed image data may be complex numbers. In that case, the complex conjugate relationship does not hold for the data in the k space. Therefore, the estimation result is incorrect and causes the image quality of the reconstructed image to deteriorate.

  In Patent Document 3, complex conjugate data is obtained from a part of k-space data, and phase correction processing is performed in a space (image space) obtained by performing inverse Fourier transform on the data to obtain estimated data. The obtained estimation data and measurement data are synthesized by returning them to the k space. For this reason, it is necessary to perform Fourier transform on the estimated data and return to the k-space again. Therefore, the number of Fourier transform processes performed in the entire estimation process is three times that in the case where the estimation process is not performed. Furthermore, when refinement is performed by the iterative process described in Patent Document 3, a larger amount of processing is required. For this reason, there is a problem that the method disclosed in Patent Document 3 is not practical in terms of processing time. (See FIG. 1B of Patent Document 3)

  Further, FIG. 1C of Patent Document 3 suggests that the estimation data and the measurement data are synthesized on the image space as an alternative embodiment, but there is little explanation in the specification, and details are unknown. It is. As shown in FIG. 1C, since the measurement data itself is used as the phase correction data, phase distortion due to the asymmetric shape can occur in the obtained phase map. For this reason, phase correction cannot be performed with high accuracy. In addition, in the method of Patent Document 3, there is an overlapping area (frequency component) in the k space in the estimation data and the measurement data. Therefore, when these image data are added, an image in which a specific frequency component is enhanced / suppressed is obtained. Therefore, there is a problem that the image quality deteriorates.

  On the other hand, since the estimation method described in Patent Document 2 performs estimation processing for each blade data, a phase map used for phase correction is created using a small number of echo signals in the blade data. Therefore, the SN ratio of the phase map is low, and the estimation processing for each blade data has low estimation accuracy. Also, the creation size of the phase map differs depending on the number of phase encodings in the blade data. Therefore, there is a problem that the estimation accuracy changes depending on the number of phase encodings in the Blade data, and the image quality is not stable.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to provide a magnetic resonance imaging apparatus capable of highly accurate estimation of an echo signal suitable for the radial sampling method.

  In order to achieve the above object, according to the first aspect of the present invention, the following magnetic resonance imaging apparatus is provided. That is, an apparatus having a signal acquisition unit that acquires a nuclear magnetic resonance signal by applying a high-frequency pulse and a gradient magnetic field to a subject, and a signal processing unit that reconstructs an image from the nuclear magnetic resonance signal. The unit acquires a plurality of nuclear magnetic resonance signal data arranged radially from the origin in the measurement space. The signal processing unit includes a filter for selecting, in the measurement space, data of a first region that includes an origin, a predetermined range from the origin, and data of a second region that includes a range farther from the origin than the first region. The first and second region nuclear magnetic resonance signal data selected by the setting unit and the filter setting unit are subjected to inverse Fourier transform to obtain first region inverse Fourier transformed data and second region inverse Fourier transformed data. Inverse Fourier transform unit, complex conjugate unit that obtains estimated data by obtaining complex conjugate data of second region inverse Fourier transformed data, and first region inverse Fourier transform from second region inverse Fourier transformed data or complex conjugate data A phase correction unit that performs phase correction by subtracting or adding the phase value of the post-data, the data after the first region inverse Fourier transform, and the second region inverse Fourier transform And an addition unit to obtain a reconstructed image by adding the data and the estimated data. When the estimated data is subjected to Fourier transform and virtually arranged in the measurement space, the estimated data after the Fourier transform does not overlap with the data in the first and second regions in the measurement space. In this manner, the data of the first and second areas are selected. As described above, by selecting the data of the first region including the center and the second region including the outside thereof, it is possible to obtain the reconstructed image by adding the estimated data in the image space. In the present invention, the meaning of “not overlapping” means not emphasizing / suppressing a part of k-space data, and is not limited to the meaning that coordinate positions in k-space do not overlap. For example, even if the coordinates of the regions overlap, if the data is filtered (selected) by changing the ratio, the k-space data is not emphasized / suppressed, and thus is included in the “non-overlapping” of the present invention. . Hereinafter, the expression “does not overlap”, “does not overlap” or “does not cause overlapping regions” is used in the present specification with the same meaning.

  For example, the filter setting unit is configured to select the data of the first and second regions so that the data of the first region and the data of the second region do not overlap in the measurement space.

  The nuclear magnetic resonance signal data acquired by the signal acquisition unit may be configured such that data is missing in a part of a region that is a predetermined distance or more away from the origin in the measurement space.

  The signal acquisition unit, for example, is configured to acquire a plurality of nuclear magnetic resonance signal data so as not to have a point-symmetric relationship with respect to the origin when arranged in a measurement space.

  Further, for example, the signal acquisition unit applies only the first half of a pulse having a symmetric function envelope in the time axis direction as a high-frequency magnetic field pulse, and acquires a signal starting from the origin of the measurement space as a nuclear magnetic resonance signal. It is possible to use a so-called ultra-short TE imaging method.

  According to the second aspect of the present invention, the following magnetic resonance imaging apparatus is provided. That is, an apparatus having a signal acquisition unit that acquires a nuclear magnetic resonance signal by applying a high-frequency pulse and a gradient magnetic field to a subject, and a signal processing unit that reconstructs an image from the nuclear magnetic resonance signal. The unit acquires a plurality of nuclear magnetic resonance signal data arranged radially from the origin in the measurement space. The signal processing unit is configured to select, in the measurement space, data of a first area that includes an origin, a predetermined range from the origin, and data of a second area that includes a range farther from the origin than the first area. The filter setting unit, the nuclear magnetic resonance signal data of the first and second regions selected by the filter setting unit, and the nuclear magnetic resonance signal data of the entire measurement space are subjected to inverse Fourier transform, respectively, and the first region is inverted. Inverse Fourier transform unit for obtaining post-Fourier transform data, second region inverse Fourier transform data and overall inverse Fourier transform data, and complex data for obtaining estimated data by obtaining complex conjugate data of the second region inverse Fourier transform data By subtracting or adding the phase value of the data after the first region inverse Fourier transform from the conjugate part and the data after the second region inverse Fourier transform or the complex conjugate data A phase correction unit that performs phase correction, and an addition unit that obtains a reconstructed image by adding the data after the entire inverse Fourier transform and the estimation data.

  The plurality of nuclear magnetic resonance signals acquired by the signal acquisition unit are divided into a plurality of groups, for example, and the plurality of nuclear magnetic resonance signals constituting each group are positioned in parallel in the measurement space, and each group is centered on the origin. The structure is arranged in a radial pattern.

  In addition, a plurality of parallel nuclear magnetic resonance signals constituting each group are configured to be positioned asymmetrically with respect to the origin, for example. In this case, it is desirable that the plurality of parallel nuclear magnetic resonance signals constituting each group are arranged so as to surround the origin with a uniform density.

  The filter setting unit outputs asymmetric nuclear magnetic resonance signal data of the second region in which no symmetric nuclear magnetic resonance signal exists around the origin in the group among a plurality of parallel nuclear magnetic resonance signals constituting each group. It is desirable to select as data.

Moreover, according to the 3rd aspect of this invention, the following image reconstruction programs are provided. That is, in the measurement space, a plurality of nuclear magnetic resonance signal data arranged radially about the origin, including the origin, data of the first area within a predetermined range from the origin, and farther from the origin than the first area Filter setting means for selecting data of the second region including the range,
Inverse Fourier transform means for performing inverse Fourier transform on the nuclear magnetic resonance signal data of the first and second regions selected by the filter setting means, respectively, and obtaining data after the first region inverse Fourier transform and data after the second region inverse Fourier transform;
Complex conjugate means for obtaining estimated data by obtaining complex conjugate data of the second domain inverse Fourier transformed data;
Phase correction means for performing phase correction by subtracting or adding the phase value of the data after the first region inverse Fourier transform from the data after the second region inverse Fourier transform or the complex conjugate data; and
Adding means for obtaining a reconstructed image by adding the data after the first region inverse Fourier transform, the data after the second region inverse Fourier transform and the estimated data;
Is a program for functioning as
When the estimated data is Fourier-transformed and virtually arranged in the measurement space, the filter setting means is configured so that the estimated data after the Fourier transform does not overlap with the data in the first and second regions in the measurement space. This is an image reconstruction program for setting the first and second areas.

Moreover, according to the 4th aspect of this invention, the following image reconstruction programs are provided. That is, in the measurement space, a plurality of nuclear magnetic resonance signal data arranged radially about the origin, including the origin, data of the first area within a predetermined range from the origin, and farther from the origin than the first area Filter setting means for selecting data of the second region including the range,
The first and second regions of the nuclear magnetic resonance signal data selected by the filter selection means and the entire nuclear magnetic resonance signal data of the measurement space are subjected to inverse Fourier transform, and the first region inverse Fourier transformed data and the second region Inverse Fourier transform means for obtaining data after inverse Fourier transform and data after overall inverse Fourier transform,
Complex conjugate means for obtaining estimated data by obtaining complex conjugate data of the second domain inverse Fourier transformed data;
Phase correction means for performing phase correction by subtracting or adding the phase value of the data after the first region inverse Fourier transform from the data after the second region inverse Fourier transform or the complex conjugate data; and
An adding means for obtaining a reconstructed image by adding the total inverse Fourier transformed data and the estimated data;
This is an image reconstruction program for functioning as

  According to the present invention, truncation artifacts and aliases appearing in a reconstructed image can be estimated with high accuracy by estimating data of a missing region in which k-space echoes are not arranged during fractional echo measurement and fractional scan measurement combined with a radial sampling method. It is possible to provide a medical image processing apparatus that can suppress mixing of components.

It is a block diagram which shows the whole structure of the MRI apparatus of 1st Embodiment. (a) And (b) It is explanatory drawing of k space data made into a process target in 1st Embodiment. It is explanatory drawing which shows the pulse sequence performed in 1st Embodiment. It is a block diagram which shows the structure of the echo signal estimation processing apparatus 301 of 1st Embodiment. It is a flowchart which shows the flow of a process of 1st Embodiment. It is a block diagram of the frequency component separation processing system 304 of the first embodiment. It is explanatory drawing of the fractional echo signal acquired in 1st Embodiment. It is explanatory drawing which shows notionally the process of Formula (9) in 1st Embodiment. (a) And (b) It is explanatory drawing of k space data made into a process target in 2nd Embodiment. It is a block diagram which shows the structure of the echo signal estimation processing apparatus 701 of 2nd Embodiment. (a) It is explanatory drawing of the data 803 for low frequency components at the time of the hybrid radial scan in 2nd Embodiment, and the data 804 for high frequency components (b). It is a block diagram which shows the structure of the k space arrangement | positioning processing system 702 in 2nd Embodiment. It is a block diagram which shows the structure of the frequency component separation processing system 704 in 2nd Embodiment. It is explanatory drawing which shows the pulse sequence performed in 3rd Embodiment. (a) And (b) It is explanatory drawing of the k space data acquired by the ultra-short TE imaging made into a 3rd embodiment. (a) k-space data (all frequency component data) of echo signals obtained by fractional echo measurement to be processed by the fourth embodiment, (b) high frequency component data, (c) low frequency component data, ( d) It is explanatory drawing of high frequency component data. It is explanatory drawing of the pulse sequence performed in 4th Embodiment. (a) And (b) It is explanatory drawing which showed two types of phase encoding patterns for Blade data collection in 4th Embodiment. It is a block diagram which shows the structure of the echo signal estimation processing apparatus 1401 of 4th Embodiment. It is a flowchart which shows the flow of the process of 4th Embodiment. It is a block diagram which shows the structure of the k space arrangement | positioning processing system 1402 in 4th Embodiment. It is a block diagram which shows the structure of the frequency component separation processing system 1404 in 4th Embodiment. It is explanatory drawing which shows notionally the process of Formula (20) in 4th Embodiment.

  Hereinafter, embodiments of the MRI apparatus of the present invention will be described in detail with reference to the accompanying drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiments of the invention, and the repetitive description thereof is omitted.

(First embodiment)
First, an overall outline of the MRI apparatus of the first embodiment will be described based on the block diagram shown in FIG. This MRI apparatus obtains a tomographic image of a subject using an NMR phenomenon. As shown in FIG. 1, the MRI apparatus includes a static magnetic field generation system 2, a gradient magnetic field generation system 3, a transmission system 5, a reception system 6, a signal processing system 7, a sequencer 4, and an operation unit 25. It is prepared for.

  The static magnetic field generation system 2 includes a static magnetic field generation source arranged around the subject 1 and generates a uniform static magnetic field in a space around the subject 1. The direction of the static magnetic field is set in the direction of the body axis in the case of the horizontal magnetic field method, and the direction orthogonal to the body axis in the case of the vertical magnetic field method. The static magnetic field generation source may be a permanent magnet system, a normal conduction system, or a superconductivity system.

  The transmission system 5 irradiates the subject 1 with a radio frequency (RF) pulse in order to cause nuclear magnetic resonance to occur in the nuclear spins of the atoms constituting the living tissue of the subject 1, and the high frequency oscillator 11 and the modulator 12 And a high-frequency amplifier 13 and a transmission-side high-frequency coil (transmission coil) 14a. The modulator 12 receives the high-frequency pulse output from the high-frequency oscillator 11, modulates the amplitude according to the timing according to the command from the sequencer 4, and outputs the modulated signal to the high-frequency amplifier 13. The high frequency amplifier 13 amplifies the high frequency pulse and then supplies the high frequency pulse to the high frequency coil 14a disposed in the vicinity of the subject 1. As a result, the subject 1 is irradiated with an RF pulse from the high-frequency coil 14a, nuclear magnetic resonance occurs in the nuclear spin of the biological tissue of the subject 1, and a nuclear magnetic resonance signal (NMR signal) is generated.

  The gradient magnetic field generating system 3 includes a gradient magnetic field coil 9 including a plurality of coils wound respectively in three axis directions of X, Y, and Z, which are coordinate systems (stationary coordinate systems) of the MRI apparatus, and each gradient magnetic field coil. And a gradient magnetic field power supply 10 to be driven. The gradient magnetic field power supply 10 applies gradient magnetic fields Gx, Gy, and Gz in the three axial directions of X, Y, and Z by supplying current to each coil in accordance with a command from the sequencer 4 described later. At the time of imaging, a slice direction gradient magnetic field pulse (Gs) is applied in a direction orthogonal to the slice plane (imaging cross section) to set a slice plane for the subject 1, and the remaining two orthogonal to the slice plane and orthogonal to each other A phase encoding direction gradient magnetic field pulse (Gp) and a frequency encoding direction gradient magnetic field pulse (Gf) are applied in one direction, and position information in each direction is encoded in the NMR signal.

  The reception system 6 includes a reception-side high-frequency coil (reception coil) 14b, a signal amplifier 15, a quadrature detector 16, and an A / D converter 17. The NMR signal from the subject 1 is detected by the high-frequency coil 14b arranged close to the subject 1, amplified by the signal amplifier 15, and then orthogonalized by the quadrature detector 16 at the timing according to the command from the sequencer 4. The A / D converter 17 converts each signal into a digital quantity and sends it to the signal processing system 7.

  The sequencer 4 operates under the control of the digital signal processing device 8 of the signal processing system 7, and sends various commands necessary for collecting the tomographic image data of the subject 1 to the transmission system 5, the gradient magnetic field generation system 3, and the reception system 6. Send to. Thereby, the RF pulse and the gradient magnetic field pulse are repeatedly applied to the subject 1 in a predetermined pulse sequence, and an echo signal, for example, is acquired as the NMR signal.

  The signal processing system 7 performs various data processing and display and storage of processing results, and includes a storage device such as an optical disk 19, a magnetic disk 18, a ROM 21, and a RAM 22, and a display 20 such as a CRT. Data from the receiving system 6 is input to the digital signal processing device 8, and the digital signal processing device 8 executes processing such as signal processing and image reconstruction, and the resulting tomographic image of the subject 1 is displayed on the display 20. The information is displayed and recorded on the magnetic disk 18 of the external storage device.

  Further, the digital signal processing device 8 includes an echo signal estimation processing device 301 as shown in FIG. The echo signal estimation processing device 301 performs a process of estimating data for a region where an echo signal in the k space is not acquired. The configuration of the echo signal estimation processing device 301 will be described in detail later.

  The operation unit 25 is used by an operator to input various control information of the MRI apparatus and control information of processing performed by the signal processing system 7, and includes a trackball or mouse 23 and a keyboard 24. The operation unit 25 is disposed close to the display 20 so that the operator can interactively control various processes of the MRI apparatus through the operation unit 25 while looking at the display 20.

  In FIG. 1, the high-frequency coil 14a and the gradient magnetic field coil 9 on the transmission side are opposed to the subject 1 in the static magnetic field space of the static magnetic field generating system 2 in which the subject 1 is inserted, If it is a horizontal magnetic field system, it is installed so as to surround the subject 1. The high-frequency coil 14b on the receiving side is installed so as to face or surround the subject 1.

  The radionuclide to be imaged by the MRI apparatus is a hydrogen nucleus (proton) that is a main constituent material of the subject as widely used in clinical practice. By imaging information on the spatial distribution of proton density and the spatial distribution of relaxation time in the excited state, the form or function of the human head, abdomen, limbs, etc. can be imaged.

  The overall operation of the MRI apparatus of the first embodiment will be described. In particular, the configuration and operation of the echo signal estimation processing device 301 will be described in detail.

In the present embodiment, the fractional echo signal is measured by a radial sampling method. The radial sampling method is a method in which a plurality of echo signals having different rotation angles with respect to the origin of the k space are arranged in a radial pattern. Data 201 and 202 in FIGS. 2A and 2B are examples in which the fractional echo signal data acquired in the present embodiment is arranged in the k space. The data 201 in FIG. 2A is an example of k-space data by radial scan, and one echo signal (Blade) to which no phase encoding is added is arranged for each rotation angle. Data 202 in FIG. 2B is an example of k-space data by hybrid radial scan, and echo signal groups (Blade) to which phase encoding is added are arranged for each rotation angle. In both the data 201 and 202, only a part of the echo signal is arranged in the region where ky <0. When the number of blades is N and the blade number is n, the k-space arrangement angle [rad] of each blade is set, for example, according to the following equation (2).

  An example of a pulse sequence for realizing the data 201 in FIG. 2 (a) or the data 202 in FIG. 2 (b) in which the k-space arrangement angle [rad] of each Blade is set by the above equation (2) is shown in FIG. Shown in The pulse sequence in FIG. 3 is an example of a radial sampling method using a gradient echo method. The horizontal axis in FIG. 3 is the time axis [s], the vertical axis is the amplitude of the RF pulse (RF), and the gradient magnetic field pulse (Gs, Gp, Gf) is the gradient magnetic field strength [T / m]. The signal (Echo) is the voltage amplitude [V] of the received echo. As can be seen from the pulse sequence of FIG. 3, in the radial sampling method, after irradiating the RF pulse 2101, the echo signal is received while simultaneously applying the phase encoding gradient magnetic field 2103 and the frequency encoding gradient magnetic field pulse 2104. At this time, the k-space arrangement angle is changed by changing the gradient magnetic field strengths IGp and IGf applied to the Gp and Gr axes at each repetition time 2106 of the pulse sequence. Specifically, the gradient magnetic field strengths IGp and IGf are set according to the following equations (3-1) to (3-2).

ここで、nはBlade番号、BladeAngle(n)は式（２）により設定される角度[rad]、IGはBandWidth[Hz]とFOV(注目視野サイズ)[m]から公知の計算方法により計算される傾斜磁場強度である。

Here, n is a blade number, BladeAngle (n) is an angle [rad] set by equation (2), and IG is calculated by a known calculation method from BandWidth [Hz] and FOV (focused field of view size) [m]. Gradient magnetic field strength.

  As described above, the sequencer 4 executes the pulse sequence shown in FIG. 3 to obtain the k-space data 201 and 202 shown in FIG. 2A or 2B.

  In the present embodiment, the echo signal estimation processing device 301 estimates data of an area where data is not acquired from the acquired k-space data.

  The configuration of the echo signal estimation processing device 301 will be described with reference to FIG. The echo signal estimation processing device 301 is arranged in the digital signal processing device 8 in FIG. As shown in FIG. 4, the echo signal estimation processing device 301 includes a k-space arrangement processing system 302, a frequency component separation processing system 304, a two-dimensional Fourier transform processing system 305, a phase correction processing system 306, a complex conjugate processing system 307, and A synthesis processing system 308 is included. The echo signal estimation processing device 301 can be constituted by, for example, an arithmetic device and a memory in which a program is stored in advance. When the arithmetic device reads and executes the program, the k space arrangement processing system 302, the frequency It functions as a component separation processing system 304, a two-dimensional Fourier transform processing system 305, a phase correction processing system 306, a complex conjugate processing system 307, and a synthesis processing system 308.

  The processing operation of each part of the echo signal estimation processing apparatus 301 of FIG. 4 will be described with reference to the flowchart of FIG. The echo signal output from the A / D converter 17 of the receiving system 6 is a complex signal F (kr, θ), and is expressed in a polar coordinate system because the radial sampling method is used. The k space arrangement processing system 302 converts the polar coordinate system data F (kr, θ) into the orthogonal coordinate system data F (kx, ky) according to the phase encoding / frequency encoding and arranges it in the RAM 22, Form (step 51). The RAM 22 stores k-space data F (kx, ky) composed of an arbitrary number of complex signals F (kr, θ).

  The frequency component separation processing system 304 reads the k-space data F (kx, ky) stored in the RAM 22 and separates it into high-frequency component data F_HF (kx, ky) and low-frequency component data F_LF (kx, ky). The result is output to the two-dimensional Fourier transform processing system 305 (steps 52 and 53). At this time, the high-frequency component data F_HF (kx, ky), the low-frequency component data F_LF (kx, ky), and the estimated data obtained by taking the phase conjugate of the high-frequency component data after phase correction in a later step, The high frequency component data F_HF (kx, ky) and the low frequency component data F_LF (kx, ky) are separated so as not to cause overlapping regions (frequency components) in the k space. In addition, the low frequency component data is preferably point-symmetric in the k space with the origin of the k space as the center.

  The two-dimensional Fourier transform processing system 305 performs high-frequency component data f_HF (x, y) and low-frequency component by inverse Fourier transforming the high-frequency component data F_HF (kx, ky) and low-frequency component data F_LF (kx, ky), respectively. Data f_LF (x, y) is obtained and output to the phase correction processing system 306 and the synthesis processing system 308 (steps 54 and 55).

  The phase correction processing system 306 subtracts the phase value (phase map) of the corresponding low frequency component data f_LF (x, y) at the x and y positions from the phase value of the high frequency component data f_HF (x, y). The high-frequency component data f_HF ′ (x, y) after phase correction is obtained and output to the complex conjugate processing system 307 (step 56). At this time, since the range of the low-frequency component data F_LF (kx, ky) is set to a point-symmetrical range with respect to the k-space origin, the phase map becomes point-symmetrical on the k-space and the phase on the image space. Since distortion does not occur, phase correction can be performed with high accuracy. In addition, in the present embodiment, since the data acquisition is performed by the radial sampling method in which the density of measurement points around the origin of the k space is high, the SN ratio of the low frequency component data F_LF (kx, ky) is high, and thus the phase The accuracy of the map is increased, and phase correction with higher accuracy can be performed.

  The complex conjugate processing system 307 calculates the complex conjugate data CC (f_HF ′ (x, y)) of the high-frequency component data f_HF ′ (x, y) after phase correction, and outputs it to the synthesis processing system 308 (step 57). . CC () represents complex conjugate data in parentheses (). Complex conjugate data CC (f_HF ′ (x, y)) is estimated data.

  The synthesis processing system 308 re-adds the input estimated data CC (f_HF ′ (x, y)), high-frequency component data f_HF (x, y), and low-frequency component data f_LF (x, y). Composition image data f_composition (x, y) is obtained (step 58). The obtained reconstructed image data f_composition (x, y) is output to the magnetic disk 18, the optical disk 19, and the display 20, and stored and displayed. Since the low-frequency component data and the high-frequency component data are separated so that the low-frequency component data, the high-frequency component data, and the estimation data do not overlap in the k-space, No suppressed component (frequency component) occurs. Therefore, the synthesis processing system 308 can obtain a phase-corrected image by a simple process of simply adding the data after inverse Fourier transform in the image space (data space after inverse Fourier transform).

上式（４−１）、（４−２）において、ｋｘ、ｋｙはｋ空間中心を原点としたときの座標である。また、式（４−２）のFilterPointsは、図７に示すようにエコー信号のｒが負の部分502の長さであり、ｋ空間中心から手前側サンプル点数である。式（４−１）、（４−２）により、ｋ空間原点を中心とした点対称な領域を選択するフィルタＬ(ｋｘ、ｋｙ）が設定される。また、このフィルタＬ(ｋｘ、ｋｙ）は、式（４−１）、（４−２）から明らかなように、FilterPointsに近づくにつれ選択されるデータの割合が徐々に減少するように設定されている。 The configuration and processing operation of the frequency component separation processing system 304 will be further described. FIG. 6 is a configuration diagram of the frequency component separation processing system 304. As shown in FIG. 6, the frequency component separation processing system 304 includes a low-pass filter amplitude transfer characteristic storage unit 401, a multiplication processing system 402, and a subtraction processing system 403 stored in the RAM 22. Amplitude transfer characteristics L (kx, ky) of the low-pass filter in the low-pass filter amplitude transfer characteristic storage unit 401 are expressed by the following equations (4-1) to (4-2).

In the above equations (4-1) and (4-2), kx and ky are coordinates when the center of the k space is the origin. Further, FilterPoints in the equation (4-2) is the length of the portion 502 where r of the echo signal is negative as shown in FIG. 7, and is the number of sample points on the near side from the center of the k space. Filters L (kx, ky) for selecting a point-symmetrical region with the k-space origin as the center are set by Expressions (4-1) and (4-2). Further, as is clear from the equations (4-1) and (4-2), the filter L (kx, ky) is set so that the ratio of data to be selected gradually decreases as it approaches FilterPoints. Yes.

上記式（５）により得られた低周波成分データＦ＿ＬＦ(ｋｘ,ｋｙ)に対してステップ５４で逆フーリエ変換を施すことで得られる低周波成分データｆ＿ＬＦ(ｘ,ｙ)は、位相補正処理系306で位相補正処理を行う際に用いる位相マップとなる。 The frequency component separation processing system 304 multiplies the amplitude transfer characteristic L (kx, ky) and k-space data F (kx, ky) by the multiplication processing system 402 so that the distance from the center of the k-space is Only data of sample points smaller than the length (FilterPoints) is selected, and low frequency component data F_LF (kx, ky) is obtained.

The low frequency component data f_LF (x, y) obtained by performing the inverse Fourier transform in step 54 on the low frequency component data F_LF (kx, ky) obtained by the above equation (5) is a phase correction processing system. This is a phase map used when the phase correction processing is performed at 306.

Next, the high frequency component data F_HF (kx, ky) is obtained by subtracting the low frequency component data F_LF (kx, ky) from the k space data F (kx, ky) by the subtraction processing system 403.

  In the equation (6), the high frequency component data F_HF (kx, ky), the low frequency component data F_LF (kx, ky), and the estimated data (complex conjugate data) synthesized by the synthesis process performed in the following equation (11). ) The high-frequency component data F_HF (kx, ky) is determined so that three regions CC (f_HF ′ (x, y)) do not have overlapping regions (frequency components) in the k space. Specifically, 0 is assigned to an area in k space (here, an area where ky ≦ 0) in which estimated data obtained by the processing of equation (10) described later is arranged. As a result, the low frequency component data and the high frequency component data can be separated so that there is no overlap and there is no component that is emphasized / suppressed when estimated data obtained in later processing is added. However, the equation (6) is an equation in the case where the target is k-space data in which a region not measured in a part of the region of ky <0 exists as shown in FIGS. is there. It goes without saying that if the k-space arrangement direction of the echo signal changes, the branching condition in equation (6) also changes.

  In the present embodiment, “does not overlap” means that a part of k-space data is not emphasized / suppressed, and is not limited to the meaning that coordinate positions in k-space do not overlap. For example, even if the coordinates of the areas overlap, the low frequency component data F_LF (kx, ky) set by the filter L (kx, ky) in equations (4-1) and (4-2) When filtering (selection) is performed so that the ratio of the selected data gradually changes, the k-space data of the region is emphasized even if the coordinates overlap with the high-frequency component data F_HF (kx, ky). / Because it is not suppressed, it is included in the “non-overlap” of the present invention. Hereinafter, the expression “does not overlap”, “does not overlap” or “does not cause overlapping regions” is used in the present specification with the same meaning.

  Further, from the equation (6), since the range of the low-frequency component data F_LF (kx, ky) is set to a point-symmetrical range with respect to the k-space origin, the phase map becomes point-symmetrical on the k-space, Since phase distortion does not occur in the image space, phase correction performed in a later process can be performed with high accuracy. In addition, in the present embodiment, since the data acquisition is performed by the radial sampling method in which the density of the measurement points around the origin of the k space is high, the SN ratio of the low frequency component data F_LF (kx, ky) is high, and thus the phase Map accuracy is high, and phase correction can be performed with higher accuracy.

  Here, since the low-pass filter amplitude transfer characteristic L (kx, ky) is used, it is possible to suppress ringing that occurs in data obtained by inverse Fourier transforming the separated low-frequency component data and high-frequency component data. If the ringing is strong, the phase map is distorted and appropriate phase correction cannot be performed. However, the low-pass filter amplitude transfer characteristics L (kx, By using the amplitude transfer characteristic of ky), ringing can be suppressed. However, the function of the low-pass filter amplitude transfer characteristic L (kx, ky) is not limited to the equations (4-1) and (4-2), and is a function that can suppress ringing of data after inverse Fourier transform. Any other function can be used. For example, a Hanning function, Gaussian function, Kaiser-Bessel function, or the like can be suitably used.

すなわち、ｋｙ＞０の領域では、低周波成分データＦ＿ＬＦ(ｋｘ,ｋｙ)と高周波成分データＦ＿ＨＦ(ｋｘ,ｋｙ)の和がｋ空間データＦ(ｋｘ,ｋｙ)に等しく、ｋｙ≦０の領域では、低周波成分データＦ＿ＬＦ(ｋｘ,ｋｙ)をフィルタ振幅伝達特性Ｌ(kx,ky)で除したものがｋ空間データＦ(ｋｘ,ｋｙ)に等しい、という関係である。低周波成分データと高周波成分データとが、重なりあわず、かつ、後の処理で求められる推定データが加算された時に強調／抑制される成分がないように、分離されていることがわかる。 The relationship among the k-space data F (kx, ky), the low-frequency component data F_LF (kx, ky), and the high-frequency component data F_HF (kx, ky) obtained by the above processing is as shown in Expression (7). Become.

That is, in the region where ky> 0, the sum of the low frequency component data F_LF (kx, ky) and the high frequency component data F_HF (kx, ky) is equal to the k space data F (kx, ky), and in the region where ky ≦ 0. The low frequency component data F_LF (kx, ky) divided by the filter amplitude transfer characteristic L (kx, ky) is equal to the k-space data F (kx, ky). It can be seen that the low-frequency component data and the high-frequency component data are separated so that they do not overlap, and there are no components that are emphasized / suppressed when estimated data obtained in later processing is added.

  Since the low-frequency component data and the high-frequency component data are separated so as not to overlap with each other and the estimated data obtained in the subsequent processing as in Expression (7), the composition processing system 308 is separated. Then, it is possible to obtain a phase-corrected image by a simple process of simply adding the data after inverse Fourier transform in the image space (data space after inverse Fourier transform). If the low frequency component data and the high frequency component data are not in the relationship of Expression (7), when the addition process is performed in the image space, an image in which a specific frequency component is emphasized or suppressed is obtained. Therefore, when the relationship of Expression (6) does not hold, the complex conjugate data CC (f_HF ′ (x, y)), the high frequency component data f_HF (x, y), the low frequency component data f_LF ( x, y) must be Fourier transformed and returned to k-space, then synthesized, corrected for data values in the region where the high-frequency component data and low-frequency component data overlap, and then inverse Fourier transformed again. , Processing becomes complicated.

  However, Expression (7) is similar to Expression (6), as shown in FIGS. 2A and 2B, k-space data in which a region not measured is present in a part of the region of ky <0. In the equation, the low-frequency component data and the high-frequency component data do not overlap with each other and are separated without overlapping with the estimation data. It goes without saying that if the k-space arrangement direction of the echo signal changes, the branching condition in equation (7) also changes.

ここで、ＩＤＦＴ[ ]は、括弧[ ]内を逆フーリエ変換処理することを示している。 Next, the processing operation of the two-dimensional Fourier transform processing system 305 in FIG. 4 will be described in detail. The two-dimensional Fourier transform processing system 305 performs inverse Fourier transform on the input low frequency component data F_LF (kx, ky) and high frequency component data F_HF (kx, ky) according to the following equations, respectively, Frequency component data f_LF (x, y) and high frequency component data f_HF (x, y) are obtained.

Here, IDFT [] indicates that an inverse Fourier transform process is performed in parentheses [].

ここで、θ＿ＨＦ(ｘ,ｙ)は、高周波成分データｆ＿ＨＦ(ｘ,ｙ)の位相値であり、θ＿ＬＦ(ｘ,ｙ)は、低周波成分データｆ＿ＬＦ(ｘ,ｙ)の位相値である。 Next, the phase correction processing system 306, the complex conjugate processing system 307, and the synthesis processing system 308 will be described in detail. The phase correction processing system 306 obtains the phase value (phase map) of the input low frequency component data f_LF (x, y) after inverse Fourier transform, and the high frequency component data f_HF (x, y) after inverse Fourier transform. Subtract from the phase value at the corresponding x, y position. Thereby, the phase correction of the high frequency component data is performed. Specifically, high-frequency component data f_HF ′ (x, y) after phase correction is obtained according to the following equation (9).

Here, θ_HF (x, y) is a phase value of the high frequency component data f_HF (x, y), and θ_LF (x, y) is a phase value of the low frequency component data f_LF (x, y).

  Since the phase correction is performed on the data after the inverse Fourier transform by the processing of Expression (9), even when the final reconstructed image data is a complex number, the data can be estimated by the phase correction and the complex conjugate.

ここで、θ＿ＨＦ'(ｘ,ｙ)は、位相補正後の高周波成分データｆ＿ＨＦ'(ｘ,ｙ)の位相値である。式（１０）から、ｋ空間上の欠落した領域の推定データCC(ｆ＿ＨＦ'(ｘ,ｙ))が算出される。この推定データをフーリエ変換してｋ空間上に戻した場合、フーリエ変換後の推定データはｋ空間上で低周波成分データＦ＿ＬＦ(ｋｘ,ｋｙ)と高周波成分データＦ＿ＨＦ(ｋｘ,ｋｙ)と重なり合わない。 The complex conjugate processing system 307 performs the calculation of the following equation (10) using f_HF ′ (x, y) calculated by the equation (9) as an input value, thereby performing high-frequency component data f_HF ′ (x, The complex conjugate data CC (f_HF ′ (x, y)) of y) is calculated.

Here, θ_HF ′ (x, y) is a phase value of the high-frequency component data f_HF ′ (x, y) after phase correction. From the equation (10), estimated data CC (f_HF ′ (x, y)) of the missing region in the k space is calculated. When the estimated data is Fourier transformed and returned to the k space, the estimated data after the Fourier transform overlaps the low frequency component data F_LF (kx, ky) and the high frequency component data F_HF (kx, ky) on the k space. Absent.

式（１１）により算出したｆ＿composition(x,y)が、最終的な再構成画像データとなる。 The synthesis processing system 308 calculates the estimated data CC (f_HF ′ (x, y)), the low frequency component data f_LF (x, y) after the inverse Fourier transform, and the high frequency component data f_HF (x) after the inverse Fourier transform. , y) is synthesized according to the following equation (11).

The f_composition (x, y) calculated by the equation (11) is the final reconstructed image data.

  FIG. 8 conceptually illustrates equation (11). As shown in FIG. 8, the estimated data CC (f_HF ′ (x, y)), the high frequency component data f_HF (x, y), and the low frequency component data f_LF (x, y) are added to each other, so that It can be seen that the missing region of the data included in the k-space data F (kx, ky) is supplemented, and f_composition (x, y) is formed as complete k-space data.

  As described above, in the present embodiment, the phase value is obtained from the low-frequency component data f_LF (x, y) after the inverse Fourier transform, and the phase value of the high-frequency component data after the inverse Fourier transform is expressed by equation (9). The complex conjugate data is obtained from the corrected high frequency component data. Thereby, good reconstructed image data can be obtained without causing a phase difference between the measurement data and the estimation data. Therefore, unlike the conventional estimation process using the complex conjugate property described in Non-Patent Document 2, it is possible to avoid a phenomenon in which the estimation result when the reconstructed image data is a complex number deteriorates the image quality. When the estimated data CC (f_HF ′ (x, y)) is Fourier transformed and returned to the k space, the estimated data after the Fourier transformation is the low frequency component data F_LF (kx, ky) on the k space. And the high frequency component data F_HF (kx, ky) do not overlap with each other, so that no frequency component that is emphasized / suppressed in the image space is generated.

  The low-frequency component data used for phase correction uses the entire central region of k-space data formed by the radial sampling method. Since the central region of the k-space data formed by the radial sampling method is a region formed by overlapping a plurality of blade data, it is the region having the highest SN ratio in the k-space data. Therefore, Patent Document 2 discloses that phase correction is performed using a single conventional blade data by creating a phase map using low-frequency region data of the entire central region of k-space data obtained by adding a plurality of blade data. It is possible to obtain a highly accurate phase map with less errors due to magnetic field distortion, noise, and the like than the above technique. Therefore, in the present embodiment in which the missing region in the k space is estimated using a highly accurate phase map, it is possible to obtain estimated data with higher accuracy than the conventional method, and using this, a highly accurate reconstructed image can be obtained. Is obtained.

  Furthermore, in this embodiment, since the estimation process can be performed by a simple process, a practical processing speed can be realized. Specifically, since the k-space data is separated into the high-frequency component data and the low-frequency component data so as to maintain the relationship of Expression (7), the synthesis processing in the synthesis processing system 308 is simply added in the image space. This makes it possible to significantly reduce the processing time during the synthesis process. On the other hand, the image forming method described in Patent Document 3 of the prior art is concerned about a decrease in practicality due to an increase in processing amount.

  In the first embodiment, after phase correction is performed on the high frequency component data in step 56 of FIG. 5, the complex conjugate is obtained in step 57 to obtain estimated data. However, the present invention is not limited to this method, and step 56 and step 57 may be interchanged to obtain a complex conjugate of high-frequency component data, and then phase correction may be performed to obtain estimated data. In this case, phase correction is performed by adding the phase of the phase map to the phase of the complex conjugate data.

(Second Embodiment)
In the first embodiment described above, as shown in FIGS. 2A and 2B, k-space data 201 in which echo signal data is missing in a part of the k-space (for example, ky <0), 202, an estimation process is performed and a reconstructed image is obtained. In the second embodiment, as shown in FIGS. 9A and 9B, the data missing portion of the fractional echo includes two echo signals. The k-space data 601 and 602 distributed so as to be sandwiched between them are subjected to estimation processing to obtain a reconstructed image. The k-space data 601 and 602 in FIGS. 9 (a) and 9 (b) are the first implementation because the missing portions of the echo signal data are distributed over the entire k-space compared to FIGS. 2 (a) and (b). There is an advantage that artifacts of the reconstructed image are less conspicuous compared to k-space data in which missing portions are collected in a specific region (for example, ky <0) as in the form.

すなわち、式（１２）に従うことにより、計測される全てのBladeデータをｋ空間の原点を中心として点対称にならないように配置できる。点対称にならないように配置することで、推定処理により得られる推定データが、計測されたエコー信号と重ならないため、推定処理後のｋ空間データの信号密度が推定処理前に比べて密になり、エイリアシングなどのアーチファクトの低減が可能となる。すなわち、式（１２）により、推定データと計測されたエコー信号のｋ空間上の位置が重ならないように配置することを意図してｋ空間配置角度を設定することができる。 The data 601 in FIG. 9A is an example of k-space data by radial scan, the data 602 in FIG. 9B is an example of k-space data by hybrid radial scan, the number of blades is N, and the blade number is When n is set, the k-space arrangement angle BladeAngle [rad] of each Blade can be obtained by setting, for example, according to the following equation (12).

That is, by following Formula (12), all the measured Blade data can be arranged so as not to be point-symmetric with respect to the origin of the k space. By arranging so that it is not point-symmetric, the estimated data obtained by the estimation process does not overlap with the measured echo signal, so the signal density of the k-space data after the estimation process is denser than before the estimation process. Artifacts such as aliasing can be reduced. That is, according to the equation (12), the k-space arrangement angle can be set with the intention of arranging the estimated data and the measured echo signal so that the positions in the k-space do not overlap.

  The specific pulse sequence is the same as the pulse sequence of FIG. 3 of the first embodiment, but the values of the gradient magnetic field strengths IGp and IGf shown in FIG. 3 are represented by BladeAngle [rad] determined by Expression (12). It is realizable by determining using a value, Formula (3-1), and Formula (3-2).

  If the k-space data 601 and 602 shown in FIGS. 9A and 9B are obtained by executing the pulse sequence determined in this way, in this embodiment, the echo signal estimation processing device 701 performs the data missing portion. The echo signal is estimated and image reconstruction is performed. The echo signal estimation processing device 701 is arranged in the digital signal processing device 8 of the MRI apparatus having the configuration shown in FIG. 1 like the echo signal estimation processing device 301 of the first embodiment. Since the other apparatus configuration is the same as that of the first embodiment, the description of the other configuration is omitted here.

  The configuration of the echo signal estimation processing device 701 will be described using FIG. As shown in FIG. 10, the echo signal estimation processing device 701 includes a k-space arrangement processing system 702, a frequency component separation processing system 704, a two-dimensional Fourier transform processing system 705, a phase correction processing system 706, a complex conjugate processing system 707, and A synthesis processing system 708 is included.

  The processing operation of this embodiment will be described with reference to FIG. The k-space arrangement processing system 702 converts the complex signal F (kr, θ) received from the A / D converter 17 into orthogonal coordinate system data in accordance with the equation (12) and the phase encoding / frequency encoding, and the RAM 22. To form k-space data. At this time, in the second embodiment, the k space arrangement processing system 702 performs the low frequency component data (F_LFpre (kx, ky)) 803 and the high frequency component data (F_HFpre (kx, ky)) shown in FIG. Two types of orthogonal coordinate system k-space data 804 are created. Therefore, two k-space data 803 and 804 are stored in the RAM 22. The frequency component separation processing system 704 reads the low-frequency component data (F_LFpre (kx, ky)) 803 and the high-frequency component data (F_HFpre (kx, ky)) 804 stored in the RAM 22 to obtain a low-pass filter. Is multiplied by the amplitude transfer characteristics of the high-pass filter and low-frequency component data F_LF (kx, ky) and high-frequency component data F_HF (kx, ky) are calculated and output to the two-dimensional Fourier transform processing system 705 To do.

  The two-dimensional Fourier transform processing system 705 performs high-frequency component data f_HF (kx, ky) and low-frequency component data f_LF obtained by performing inverse Fourier transform on the high-frequency component data F_HF (kx, ky) and the low-frequency component data F_LF (kx, ky), respectively. (kx, ky) is calculated and output to the phase correction processing system 706 and the synthesis processing system 708. The phase correction processing system 706 corrects the phase by subtracting the phase value of the low-frequency component data f_LF (kx, ky) from the high-frequency component data f_HF (kx, ky), and the obtained phase correction data f_HF ′ (kx, ky) , ky) is output to the complex conjugate processing system 707. The complex conjugate processing system 707 calculates the complex conjugate data CC (f_HF ′ (kx, ky)) of the phase correction data f_HF ′ (kx, ky), and outputs it to the synthesis processing system 708. The complex conjugate data CC (f_HF ′ (kx, ky)) is estimated data of the data missing part in FIGS. 9A and 9B.

  The synthesis processing system 708 performs addition processing on the complex conjugate data CC (f_HF ′ (kx, ky)), high-frequency component data f_HF (kx, ky), and low-frequency component data f_LF (kx, ky), which are estimation data. Thus, reconstructed image data f_composition (x, y) is obtained and output to the magnetic disk 18, the optical disk 19, and the display 20, and stored and displayed.

ただし、ｋｘ＝ｋｒ×ｃｏｓ(θ)、ｋｙ＝ｋｒ×ｓｉｎ(θ)である。
また、式（１３−１）においてSymmetricRangeは、図７中の部分502の長さを表す。 The configuration and processing operation of the k space arrangement processing system 702 will be described in more detail with reference to FIG. As shown in FIG. 12, the k space arrangement processing system 702 includes a low frequency component data creation processing system 801 and a high frequency component data creation processing system 802. In the present embodiment, low frequency component data and high frequency component data are separated at the stage of the k space arrangement processing. For this purpose, two k-space regions are prepared and an arrangement process is performed. The low frequency component data creation processing system 801 extracts only the signal in the range of the portion 503 from the fractional echo signal shown in FIG. F_LFpre (kx, ky)) 803 is formed. The high-frequency component data creation processing system 802 extracts only the signal in the range of the portion 504 from the fractional echo signal shown in FIG. 7 and arranges it in the area on the RAM 22, whereby k-space data (F_HFpre (kx , ky)) 804. FIG. 11 shows an example of the k space 803 based on the low frequency component data during the hybrid radial scan and the k space 804 based on the high frequency component data during the hybrid radial scan. Specifically, these k-space forming processes follow equations (13-1) to (13-2).

However, kx = kr × cos (θ) and ky = kr × sin (θ).
Further, in formula (13-1), SymmetricRange represents the length of the portion 502 in FIG.

  In the present embodiment, the purpose of preparing high-frequency component data and low-frequency component data as k-space data is to obtain estimated data and measurement data that do not overlap frequency bands in the subsequent processing. . By not overlapping the frequency bands of the estimated data and the measured data, the synthesis data 708 can simply add the estimated data and the measured data.

Next, the processing operation of the frequency component separation processing system 704 will be described in detail with reference to FIG. As shown in FIG. 13, the frequency component separation processing system 704 includes a low-pass filter amplitude transfer characteristic storage unit 1001 and a high-pass filter amplitude transfer characteristic storage unit 1002 stored in the RAM 22, and two multiplication processes. System 1003. Since the k-space arrangement processing system 702 prepares two types of data, low frequency component data (F_LFpre (kx, ky)) 803 and high frequency component data (F_HFpre (kx, ky)) 804, the frequency component The separation processing system 704 performs processing for multiplying these two types of data by the amplitude transfer characteristics of the filter. The amplitude transfer characteristic L (kx, ky) of the low-pass filter is set by the equations (4-1) to (4-2) of the first embodiment. The amplitude transfer characteristic H (kx, ky) of the high-pass filter is set by the following equation (14).

By multiplying the amplitude transfer characteristic L (kx, ky) of the low-pass filter and the low-frequency component data (F_LFpre (kx, ky)) 803 as shown in Equation (15), the low-frequency component data F_LF ( kx, ky).

  By performing inverse Fourier transform on the low-frequency component data F_LF (kx, ky) obtained here, f_LF (kx, ky) serving as a phase map when the phase correction processing system 306 performs the phase correction processing is obtained. .

Next, the high-frequency component data F_HF (kx, ky) is multiplied by the amplitude transfer characteristic H (kx, ky) of the high-pass filter and the high-frequency component data (F_HFpre (kx, ky)) 804 as shown in Expression (16). , ky).

式（１７）が成り立つことにより、低周波成分データと高周波成分データとが重なりあわず、かつ、後の処理で求められる推定データが加算されたときに強調／抑制される成分（周波数成分）が生じないように分離されている関係にあるため、合成処理系308では、逆フーリエ変換後のデータを画像空間（逆フーリエ変換後のデータ空間）において加算するだけの簡単な処理で位相補正された画像を得ることができる。 Also in the second embodiment, the low-frequency component data F_LF (kx, ky) and the high-frequency component data F_HF (kx, ky) have the same relational expression as the expression (7) in the first embodiment. Specifically, the low-frequency component data and the high-frequency component data are separated without overlapping with the estimated data. Since the branching condition is complicated in the orthogonal coordinate system, the relational expression is expressed as shown in Expression (17) in the polar coordinate system.

By satisfying the expression (17), the low-frequency component data and the high-frequency component data do not overlap, and the component (frequency component) that is emphasized / suppressed when the estimated data obtained in the subsequent processing is added. In the composition processing system 308, the phase correction is performed by a simple process of adding the data after the inverse Fourier transform in the image space (the data space after the inverse Fourier transform) because the relation is separated so as not to occur. An image can be obtained.

  The operations of the two-dimensional Fourier transform processing system 705, the phase correction processing system 706, the complex conjugate processing system 707, and the synthesis processing system 708 are as described above. That is, the phase correction is performed by subtracting the phase value of the low-frequency component data f_LF (x, y) from the high-frequency component data f_HF (x, y) after the two-dimensional Fourier transform processing, and the obtained phase correction data f_HF ′. The complex conjugate data CC (f_HF ′ (x, y)) of (x, y) is calculated. The synthesis processing system 708 performs addition processing on the complex conjugate data CC (f_HF ′ (x, y)), the high frequency component data f_HF (x, y), and the low frequency component data f_LF (x, y), which are estimation data. Thus, reconstructed image data f_composition (x, y) can be obtained.

  As described above, in the second embodiment, the k-space data in which the data missing portion of the fractional echo is dispersedly arranged so as to be sandwiched between two echo signals is shown in FIGS. 2 (a) and 2 (b). Thus, the estimated data of the missing part is calculated, and a reconstructed image can be obtained. Since the missing portion of the echo signal data is distributed over the entire k space as compared with FIGS. 2 (a) and 2 (b), the missing portion is a specific region (for example, ky <0) as in the first embodiment. Compared with the k-space data that is gathered together, artifacts of the reconstructed image can be suppressed.

  In the second embodiment, since the k-space arrangement angle of the echo signal is set so as to satisfy the above equation (12), all the measured echo signal data of the Blade is centered on the origin of the k-space. It is not point-symmetric and the estimated data obtained by the estimation process does not overlap with the measured echo signal. As a result, the signal density of the k-space data after the estimation process becomes denser than that before the estimation process, and artifacts such as aliasing can be reduced.

  Further, in the second embodiment, the k-space data (FIGS. 9 (a) and 9 (b)) having a complicated pattern in which the missing data portion of the echo signal is distributed so as to be sandwiched between the two echo signals. However, as shown in FIG. 11, the frequency components are distinguished and k-space arrangement is performed, so that estimation processing and synthesis processing on the space after the inverse Fourier transform can be performed.

  Also in the second embodiment, similarly to the first embodiment, the phase correction is performed after the inverse Fourier transform to obtain complex conjugate data, so that estimation data can be obtained with high accuracy even when the reconstructed image is a complex image. The data used for phase correction is an area formed by adding a plurality of Blade data centered on k-space, so that the SN ratio is high, and accurate correction can be performed.

  As for the processing amount, compared to the first embodiment, the processing amount increases by preparing two types of k-space data for low-frequency components and high-frequency components, but the inverse Fourier that requires the most processing time. Since the conversion process is the same as that of the first embodiment, the increase rate of the entire processing amount is not large and can be put into practical use.

(Third embodiment)
The MRI apparatus of the third embodiment acquires echo signal data by a pulse sequence called ultra-short TE imaging (UTE imaging), obtains estimated data for a data missing portion, and generates a reconstructed image. The device configuration is the same as in the first embodiment.

  The UTE imaging method is a known imaging method, and is disclosed in, for example, US Pat. No. 5,025,216 and US Pat. No. 5,515,0053. In the present embodiment, an echo signal is acquired by the pulse sequence of FIG. In FIG. 14, the horizontal axis is the time axis [s], and the vertical axis is the amplitude for the RF pulse, the gradient magnetic field strength [T / m] for the gradient magnetic fields Gs, Gp, and Gf, and the echo signal (Echo ) Is the voltage amplitude [V] of the received echo. This pulse sequence excites spins by irradiating a half RF pulse 2201 of the first half of an RF pulse with an envelope of a symmetric function (for example, sinc function) at the same time as the slice selective gradient magnetic field Gs, and uses a phase gradient magnetic field. Instead, the echo signal is measured from the rise of the gradient magnetic fields Gs and Gf. Thereby, a signal can be measured in a very short time (TE) from spin excitation. By shortening TE, it is possible to image tissues having a short transverse relaxation time T2, such as cortical bone, Achilles tendon, and ligament.

  The echo signal 2205 obtained by excitation with the half RF pulse 2201 is measurement data from one side with respect to the origin when considering the slice axis in the k space. As shown in FIG. 14, the echo signal 2205 has zero sample points in the section indicated by the portion 502 in FIG. 7 in the k space. For this reason, in ultrashort TE imaging, as shown in FIG. 14, the measurement is performed twice with different polarities of the slice selection gradient magnetic field Gs applied together with the half RF pulse, and the echo signal obtained by the two measurements is complex. By adding, a signal equivalent to that obtained when a full RF pulse is used is obtained. In this embodiment, since the radial sampling method is used, the k-space arrangement angle is changed by changing the gradient magnetic field strengths IGp and IGf of the gradient magnetic field Gp2203 and the gradient magnetic field pulse 2204 every 2206. Specifically, gradient magnetic field strengths IGp and IGf are determined based on Blade Angle [rad] determined by Expression (12) of the second embodiment and Expressions (3-1) to (3-2) of the first embodiment. By setting, k-space data 1101 and 1102 having the arrangement shown in FIG. 15 (a) or (b) is obtained.

The digital signal processing device 8 of the MRI apparatus of this embodiment includes an echo signal estimation processing device 301 as in the configuration of FIG. 1, and the k-space data 1101 and 1102 of FIGS. 15 (a) and 15 (b). Image reconstruction is performed by estimating data for a region from which no echo signal has been acquired. The processing operation is almost the same as that of the first embodiment, but the difference from the first embodiment is the processing operation of the low-pass filter amplitude transfer characteristic storage unit 401 stored in the RAM 22 of FIG. In order to explain this, the formulas (4-1) and (4-3) shown in the first embodiment will be described again.

ここで、Δｋはｋ空間のサンプルピッチであり、ＮはBlade数である。すなわち式（１８）は、ｋ空間上に配置される各Bladeデータ間のデータ点間隔がｋ空間のサンプルピッチ以下になる範囲をFilterpointsで指定する式である。式（１８）に従って設定することにより、逆フーリエ変換後の低周波成分データｆ＿ＬＦ（ｘ、ｙ）にエイリアシングを発生させず、適切な位相マップを得ることができる。 In the first embodiment, the Filterpoints of the above equation (4-2) is defined as the length indicated by the portion 502 in FIG. 7, but in the third embodiment, the number of sample points in the portion 502 is zero in the echo signal. It is possible to specify the length of Filterpoints to an arbitrary length and execute the processing. For example, the following equation (18) is defined as an example.

Here, Δk is a sample pitch in k-space, and N is the Blade number. That is, Expression (18) is an expression for designating a range in which the data point interval between each Blade data arranged in the k space is equal to or smaller than the sample pitch of the k space by Filterpoints. By setting according to Expression (18), it is possible to obtain an appropriate phase map without causing aliasing in the low frequency component data f_LF (x, y) after the inverse Fourier transform.

  Thus, in this embodiment, since Filterpoints can be arbitrarily determined, the number of points in the section indicated by the portion 502 in FIG. 7 of the echo signal is zero as in the ultrashort TE imaging method to which the radial sampling method is applied. Even in this case, estimation processing can be performed. In the Cartesian sampling method, when the number of points in the section indicated by 502 in FIG. 7 is zero, the frequency band of the phase map is biased and an appropriate estimation result cannot be obtained. By using the sampling method and the Blade arrangement method shown in Expression (10), a phase map that does not cause a large deviation in the frequency band is obtained, and the estimation process can be performed.

(Fourth embodiment)
The MRI apparatuses of the first to third embodiments described above are configured to perform echo signal estimation processing at the time of fractional echo measurement, but the MRI apparatus of the present embodiment is configured to perform echo signal estimation processing at the time of fractional scan measurement. I do. In the present embodiment, a part of the echo signal in the blade is deleted so that the signal density is not biased in the central part of the k-space data to be formed. Thereby, it is possible to perform the estimation process with high accuracy using the data in the central region.

  In the present embodiment, as shown in the k-space data 1201 in FIG. 16A, fractional scan measurement is performed to reduce the number of echo signals in each blade of the hybrid radial scan. A pulse sequence will be described with reference to FIG. The pulse sequence in FIG. 17 is an example of a radial sampling method using a gradient echo method. The horizontal axis in FIG. 17 is the time axis [s], and the vertical axis is the amplitude for the RF pulse, the gradient magnetic field strength [T / m] for the gradient magnetic field pulses Gs, Gp, and Gf, and the echo signal (Echo ) Is the voltage amplitude [V] of the received echo. As shown in FIG. 17, in the hybrid radial scan, after irradiating an RF pulse 2301, gradient magnetic fields 2307 and 2308 or gradient magnetic fields 2309 and 2310 for phase encoding are applied to the Gs axis and Gp axis. The gradient magnetic fields 2307 and 2308 and the gradient magnetic fields 2309 and 2310 are obtained by reversing the signs of the gradient magnetic field strengths, and realize phase encoding in the phase encoding patterns 1301 and 1302 of each blade as shown in FIG. To be applied. Thereafter, an echo signal 2305 is obtained by applying a phase encoding gradient magnetic field 2303 and a frequency encoding gradient magnetic field pulse 2304.

  The k-space arrangement angle is changed by changing the gradient magnetic field strengths IGp and IGf of the gradient magnetic field pulses 2303 and 2304 (Gp and Gr) for each repetition 2306 and 2311 of the pulse sequence. Specifically, the k-space arrangement angle BladeAngle of Blade data is set according to Expression (2) of the first embodiment, and the gradient magnetic field strengths IGp and IGf are set according to Expressions (3-1) to (3-2). Thus, an echo signal arranged in k-space data as shown in FIG. 16 can be acquired.

  The k-space data 1201 in FIG. 16 has a configuration in which two types of phase encoding patterns 1301 and 1302 shown in FIGS. 18A and 18B are alternately arranged for each k-space arrangement angle. 18A and 18B, the solid line arrow 1303 represents the echo signal to be measured, and the dotted line arrow 1304 represents the echo signal calculated by the estimation process. The phase encoding pattern 1301 in FIG. 13 is a pattern in which the echo signal to be measured is biased toward negative phase encoding, and the phase encoding pattern 1302 is a pattern in which the echo signal to be measured is biased toward positive phase encoding. In the present embodiment, the phase encode patterns 1301 and 1302 are alternately performed for each k-space arrangement angle BladeAngle to perform measurement. As a result, the k-space data 1201 in FIG.

  As described above, the center portion of the k-space data 1201 formed by the phase encoding patterns 1301 and 1302 biased to the plus and minus sides alternately for each k-space placement angle so as to surround the origin. Since the echo signals are arranged side by side, the signal density is not biased. As a result, there is no signal density deviation, that is, no distortion, in the central portion of the k-space data that is the phase map used in the estimation process.

  The apparatus configuration of the MRI apparatus is the same as that of the first embodiment, but differs from the first embodiment in that an estimated echo signal estimation processing apparatus 1401 shown in FIG. 19 is provided inside the digital signal processing apparatus 8 of FIG. It is a point that is arranged. As shown in FIG. 19, the echo signal estimation processing device 1401 includes a k-space arrangement processing system 1402, a frequency component separation processing system 1404, a two-dimensional Fourier transform processing system 1405, a phase correction processing system 1406, a complex conjugate processing system 1407, and A synthesis processing system 1408 is included.

  The operation of each unit in FIG. 19 will be described using the flow in FIG. The k-space arrangement processing system 1402 converts the complex signal F (kr, θ) received from the A / D converter 17 into orthogonal coordinate system data according to the above equation (12) and the phase encoding / frequency encoding. Are arranged in the RAM 22 to form a k-space. At this time, as the orthogonal coordinate system data, two types of data, F_OF (kx, ky) 1201 for all frequency components in FIG. 16 (a) and F_HFpre (kx, ky) 1202 for high frequency components in FIG. 16 (b), are created. (Steps 71 and 72). The high-frequency component data F_HFpre (kx, ky) 1202 is an echo signal in which there is no echo signal having a symmetric phase encode amount centered on the phase encode amount 0 in the phase encode patterns 1301 and 1302 (echo signal in the high frequency non-target region) ) 1305 is created by selecting from echo signals for each measured blade. This process will be described in detail later.

  The frequency component separation processing system 1404 reads the all-frequency component data F_OF (kx, ky) 1201 and the high-frequency component data F_HFpre (kx, ky) 1202 stored in the RAM 22, and displays the low-frequency component in FIG. Data F_LF (kx, ky) 1203 and high-frequency component data F_HF (kx, ky) 1204 in FIG. 16 (d) are calculated and output to the two-dimensional Fourier transform processing system 1405 (steps 73 and 74). The calculation method will be described later in detail. At the same time, the frequency component separation processing system 1404 outputs the entire frequency component data F_OF (kx, ky) 1201 to the two-dimensional Fourier transform processing system 1405 as it is.

  The two-dimensional Fourier transform processing system 1405 performs inverse Fourier transform on all frequency component data F_OF (kx, ky) 1201, high frequency component data F_HF (kx, ky) 1204, and low frequency component data F_LF (kx, ky) 1203. All frequency component data f_OF (x, y), high frequency component data f_HF (x, y), and low frequency component data f_LF (x, y) are calculated (steps 75, 76, 77).

  The total frequency component data f_OF (x, y) after the inverse Fourier transform is input to the synthesis processing system 1408, and the high frequency component data f_HF (x, y) and the low frequency component data f_LF (x, y) are subjected to phase correction processing. Input to the system 1406. The phase correction processing system 1406 obtains data f_HF ′ (x, y) after phase correction by subtracting the phase value of the low frequency component data f_LF (x, y) from the high frequency component data f_HF (x, y). And output to the complex conjugate processing system 1407 (step 78). The complex conjugate processing system 1407 calculates the complex conjugate data CC (f_HF ′ (x, y)) of f_HF ′ (x, y) and outputs it to the synthesis processing system 1408 (step 79). This CC (f_HF ′ (x, y)) is estimated data. The composition processing system 1408 obtains reconstructed image data f_composition (x, y) by adding the input estimation data CC (f_HF ′ (x, y)) and the total frequency component data f_OF (x, y). The data is output to the magnetic disk 18, the optical disk 19, and the display 20 (step 80).

  Here, the processing operation of the k space arrangement processing system 1402 will be described in more detail. FIG. 21 is a configuration diagram of the k space arrangement processing system 1402. As shown in FIG. 21, the k-space arrangement processing system 1402 includes an all-frequency component data creation processing system 1501 and a high-frequency component data creation processing system 1502. In the present embodiment, two k-space regions 1503 and 1504 are prepared in the RAM 22 and arranged in order to separate all frequency component and high-frequency component data at the stage of k-space arrangement processing.

  The data generation processing system 1501 for all frequency components extracts all the fractional scan signals (echo signals 1305 and 1306 in FIG. 18) acquired by the phase encode patterns 1301 and 1302 of each blade shown in FIG. By arranging in 1503, all frequency component data F_OF (kx, ky) 1201 is formed (step 71 described above).

  The high frequency component data creation processing system 1502 extracts only the echo signal 1305 of the high frequency asymmetric region from the echo signals of the phase encode patterns 1301 and 1302 of each blade shown in FIG. And high-frequency component data F_HFpre (kx, ky) 1202 is formed (step 72 described above). The high-frequency asymmetric region is a region where there is no echo signal having a phase encode amount that is symmetric about the phase encode amount 0. By calculating the complex conjugate of the echo signal 1305 in this high-frequency asymmetric region, an unmeasured echo signal 1304 that is symmetric with respect to the echo signal 1305 can be obtained by estimation. Therefore, here, only the signal of the echo signal 1305 in the high-frequency asymmetric region is extracted, and the high-frequency component data F_HFpre (kx, ky) 1202 is formed.

  Note that the echo signal 1306 is an echo signal in a region where there is an echo signal having a symmetric phase encoding amount centered on the phase encoding amount 0.

  As described above, the purpose of preparing two types of k-space data is to prevent the frequency bands of the estimation data obtained in step 79 and the measurement data from overlapping. By not overlapping the frequency band of the estimation data and the measurement data, it is possible to simply add the estimation data and the measurement data in the synthesis processing system 1408 as described in the equation (7) shown in the first embodiment. Become.

  Next, the configuration and processing operation of the frequency component separation processing system 1404 will be described in more detail with reference to FIG. The frequency component separation processing system 1404 includes a low-pass filter amplitude transfer characteristic storage unit 1601, a high-pass filter amplitude transfer characteristic storage unit 1602, and a multiplication processing system 1603 stored in the RAM 22. The amplitude transfer characteristic L (kx, ky) of the low-pass filter is set by, for example, the equations (4-1) to (4-2) and the equation (18) of the first embodiment. That is, in the fourth embodiment, an arbitrary length can be set as Filterpoints.

ここで得られた低周波成分データＦ＿ＬＦ(kx,ky)は、ステップ76で逆フーリエ変換を施され、位相マップとなる。位相マップは、位相補正処理系1406で位相補正処理を行う際に用いられる。 The following equation (19) is used to multiply the low-pass filter amplitude transfer characteristic L (kx, ky) by the total frequency component data F_OF (kx, ky) 1201 to obtain the low frequency component data F_LF (kx, ky). Obtain (step 73 above). An example of the low frequency component data F_LF (kx, ky) is shown in FIG.

The low frequency component data F_LF (kx, ky) obtained here is subjected to inverse Fourier transform in step 76, and becomes a phase map. The phase map is used when the phase correction processing system 1406 performs phase correction processing.

PhaseLengthは、図１８におけるｋ空間上の位相エンコード方向の距離（Blade幅）1307である。式（２０）は、後に合成される推定データと計測データの周波数帯域を重複させないことを意図している。 On the other hand, the amplitude transfer characteristic H (kx, ky) of the high-pass filter is set by the following equation (20), for example.

PhaseLength is a distance (Blade width) 1307 in the phase encoding direction on the k space in FIG. Expression (20) is intended not to overlap the frequency band of estimated data and measurement data to be synthesized later.

  Regarding the fourth embodiment, the low frequency component data F_LF (kx, ky) is not added to other data in the synthesis processing system 1408 in step 80 as is apparent from FIG. For this reason, in this embodiment, unlike the formula (14) of the second embodiment, there is no need to establish a non-overlapping relationship between L (kx, ky) and H (kx, ky). That is, L (kx, ky) and H (kx, ky) can be arbitrarily set individually.

高周波成分データＦ＿ＨＦ(kx,ky)の一例が、図１６(d)のデータ1204である。 Next, the amplitude transfer characteristic H (kx, ky) of the high-pass filter and the high-frequency component data F_HFpre (kx, ky) 1202 are multiplied according to the following equation (21) to obtain the high-frequency component data F_HF (kx, ky). Extract (step 74 above).

An example of the high frequency component data F_HF (kx, ky) is the data 1204 in FIG.

式（２２）により算出したｆ＿composition(x,y)が、最終的な再構成画像データとなる。図２３は、式（２２）を概念的に図説したものである。図２３に示すように、複素共役データCC（ｆ＿ＨＦ'(x,y)）と、逆フーリエ変換後の全周波成分データｆ＿ＯＦ(x,y)とが加算されることで、元々のｋ空間データＦ(kx,ky)のデータの欠落領域が補われ、完全なｋ空間データが形成されることがわかる。 The operations of the two-dimensional Fourier transform processing system 1405, the phase correction processing system 1406, and the complex conjugate processing system 1407 are as described above. The synthesis processing system 1408 synthesizes complex conjugate data CC (f_HF ′ (x, y)), which is estimated data, and all frequency component data f_OF (x, y) after inverse Fourier transform according to the following equation (22).

F_composition (x, y) calculated by the equation (22) is the final reconstructed image data. FIG. 23 conceptually illustrates equation (22). As shown in FIG. 23, the original k-space data is obtained by adding the complex conjugate data CC (f_HF ′ (x, y)) and the total frequency component data f_OF (x, y) after the inverse Fourier transform. It can be seen that the missing area of F (kx, ky) data is compensated for and complete k-space data is formed.

  Even when a fractional scan is applied to the radial sampling method as in the fourth embodiment, estimated data can be obtained by phase correction and complex conjugate processing after inverse Fourier transform. In the fourth embodiment, data in the central region is used by deleting some echo signals in the blade so that the signal density is not biased in the central portion of the k-space data during the fractional scan. Thus, the estimated data can be obtained with high accuracy. Thereby, a reconstructed image with few artifacts can be obtained.

  In the fourth embodiment, a plurality of echo signals are acquired for each blade as shown in FIGS. 16 (a) and 18 (a), (b), but the present invention is not limited to this, The configuration may be such that one echo signal is acquired for each blade. In this case, the phase encoding amount is set in FIGS. 18A and 18B so that the acquired echo signal does not pass through the origin of the k space.

  In addition, this invention is not limited to the 1st-4th embodiment mentioned above. For example, the estimation of the echo signal can also be realized by the same procedure for the measurement combining the fractional echo measurement and the fractional scan measurement.

  Further, in the above embodiment, by satisfying the relationship of Expression (7) and Expression (17), it is possible to add estimated data and acquired data that do not cause overlapping of frequency components as image data. Data can be synthesized by simple image addition processing. However, the present invention is not limited to this configuration, and is set so as not to satisfy the relationship of Expression (7) or Expression (17), and is output as an image in which a specific frequency component is emphasized or suppressed. It can also be designed.

  In each embodiment, the processing operation as described above is performed, but a part of the processing may be omitted.

  Further, the present invention is not limited to a two-dimensional image, and can be applied to a three-dimensional image. It is also possible to implement different estimation methods for each dimension.

DESCRIPTION OF SYMBOLS 1 ... Subject, 2 ... Static magnetic field generation system, 3 ... Gradient magnetic field generation system, 4 ... Sequencer, 5 ... Transmission system, 6 ... Reception system, 7 ... Signal processing system, 8 ... Central processing unit (CPU), 9 ... Gradient magnetic field coil, 10 Gradient magnetic field power source, 11 High frequency transmitter, 12 Modulator, 13 High frequency amplifier, 14 a High frequency coil (transmitting coil), 14 b High frequency coil (receiving coil), 15 Signal amplifier, 16 ... Quadrature phase detector, 17 ... A / D converter, 18 ... Magnetic disk, 19 ... Optical disk, 20 ... Display, 21 ... ROM, 22 ... RAM, 23 ... Trackball or mouse, 24 ... Keyboard, 25 ... Operation unit , 301, 701, 1401... Echo signal estimation processing device

Claims (14)

A signal acquisition unit that applies a high-frequency pulse and a gradient magnetic field to the subject to acquire a nuclear magnetic resonance signal, and a signal processing unit that reconstructs an image from the nuclear magnetic resonance signal;
The signal acquisition unit acquires a plurality of nuclear magnetic resonance signal data arranged radially around the origin in the measurement space,
The signal processing unit selects, in the measurement space, data of a first area that includes an origin, a predetermined range from the origin, and data of a second area that includes a range farther from the origin than the first area. And a first region inverse Fourier transform data and a second region inverse Fourier transform by performing inverse Fourier transform on each of the first and second region nuclear magnetic resonance signal data selected by the filter setting unit From an inverse Fourier transform unit for obtaining post-data, a complex conjugate unit for obtaining estimation data by obtaining complex conjugate data of the data after the second region inverse Fourier transform, and data after the second region inverse Fourier transform or complex conjugate data A phase correction unit for performing phase correction by subtracting or adding a phase value of the data after the first region inverse Fourier transform, and the first region inverse Fourier transform And an addition unit to obtain a reconstructed image by adding the data and a second region inverse Fourier transformed data and the estimated data,
The filter setting unit, when the estimated data is Fourier-transformed and virtually arranged in the measurement space, the estimated data after the Fourier transform overlaps with the data of the first and second regions in the measurement space. A magnetic resonance imaging apparatus, wherein data of the first and second regions are selected so as not to match.   The magnetic resonance imaging apparatus according to claim 1, wherein the phase correction unit performs phase correction by subtracting a phase value of the data after the first region inverse Fourier transform from the data after the second region inverse Fourier transform, The magnetic resonance imaging apparatus, wherein the complex conjugate unit obtains complex conjugate data of the two-region inverse Fourier transformed data after phase correction.   2. The magnetic resonance imaging apparatus according to claim 1, wherein the filter setting unit includes the first and second filters so that data in the first region and data in the second region do not overlap in the measurement space. A magnetic resonance imaging apparatus, wherein each region data is selected.   2. The magnetic resonance imaging apparatus according to claim 1, wherein the nuclear magnetic resonance signal data acquired by the signal acquisition unit is missing in a part of the measurement space at a predetermined distance or more from the origin. A magnetic resonance imaging apparatus.   2. The magnetic resonance imaging apparatus according to claim 1, wherein when the signal acquisition unit is arranged in the measurement space, the plurality of nuclear magnetic resonance signal data are not in a point-symmetric relationship with respect to the origin. To obtain a magnetic resonance imaging apparatus.   2. The magnetic resonance imaging apparatus according to claim 1, wherein the signal acquisition unit applies only a first half of a pulse having a symmetric function as an envelope in a time axis direction as a high-frequency magnetic field pulse, and uses the measurement space as the nuclear magnetic resonance signal. A magnetic resonance imaging apparatus that acquires a signal starting from the origin of the magnetic resonance imaging apparatus. A signal acquisition unit that applies a high-frequency pulse and a gradient magnetic field to the subject to acquire a nuclear magnetic resonance signal, and a signal processing unit that reconstructs an image from the nuclear magnetic resonance signal;
The signal acquisition unit acquires a plurality of nuclear magnetic resonance signal data arranged radially around the origin in the measurement space,
The signal processing unit selects, in the measurement space, data of a first area that includes an origin, a predetermined range from the origin, and data of a second area that includes a range farther from the origin than the first area. The first and second regions selected by the filter setting unit, and the nuclear magnetic resonance signal data of the entire measurement space are subjected to inverse Fourier transform, respectively. Inverse Fourier transform unit for obtaining data after region inverse Fourier transform, data after second region inverse Fourier transform, and data after overall inverse Fourier transform, and estimated data by obtaining complex conjugate data of the data after the second region inverse Fourier transform A complex conjugate unit for obtaining the subtracted phase value of the data after the first region inverse Fourier transform from the data after the second region inverse Fourier transform or the complex conjugate data, or A phase correction unit that performs phase correction by calculation to, magnetic resonance imaging apparatus which comprises an addition unit to obtain a reconstructed image by adding the a whole inverse Fourier transformed data and the estimated data.   The magnetic resonance imaging apparatus according to claim 7, wherein the phase correction unit performs phase correction by subtracting a phase value of the data after the first region inverse Fourier transform from the data after the second region inverse Fourier transform, The magnetic resonance imaging apparatus, wherein the complex conjugate unit obtains complex conjugate data of the data after the two-region inverse Fourier transform after phase correction.   8. The magnetic resonance imaging apparatus according to claim 7, wherein the plurality of nuclear magnetic resonance signals acquired by the signal acquisition unit are divided into a plurality of groups, and the plurality of nuclear magnetic resonance signals constituting each group are included in the measurement space. The magnetic resonance imaging apparatus is characterized in that the groups are arranged in parallel, and each group is arranged radially around the origin.   The magnetic resonance imaging apparatus according to claim 9, wherein the plurality of parallel nuclear magnetic resonance signals constituting each group are positioned asymmetrically with respect to the origin.   The magnetic resonance imaging apparatus according to claim 10, wherein the plurality of parallel nuclear magnetic resonance signals constituting each group are arranged so as to surround the origin at a uniform density. apparatus.   12. The magnetic resonance imaging apparatus according to claim 11, wherein the filter setting unit includes a plurality of parallel nuclear magnetic resonance signals constituting each group, and a symmetric nuclear magnetic resonance signal centered on the origin in the group. A non-existent asymmetric nuclear magnetic resonance signal data is selected as data of the second region. Arithmetic unit
In the measurement space, a plurality of nuclear magnetic resonance signal data arranged radially from the origin, including the origin, data of the first area within a predetermined range from the origin, and farther from the origin than the first area Filter setting means for selecting data in the second region including the range;
Inverse Fourier transform for obtaining the first region inverse Fourier transform data and the second region inverse Fourier transform data by performing inverse Fourier transform on the nuclear magnetic resonance signal data of the first and second regions selected by the filter setting means, respectively. means,
Complex conjugate means for obtaining estimated data by obtaining complex conjugate data of the second region inverse Fourier transformed data;
Phase correction means for performing phase correction by subtracting or adding the phase value of the data after the first region inverse Fourier transform from the data after the second region inverse Fourier transform or the complex conjugate data; and
Adding means for obtaining a reconstructed image by adding the data after the first region inverse Fourier transform, the data after the second region inverse Fourier transform, and the estimated data;
Is a program for functioning as
In the measurement space, when the estimated data is virtually arranged in the measurement space by the Fourier transform, the filter setting means overlaps the estimated data after the Fourier transform with the data in the first and second regions in the measurement space. An image reconstruction program characterized by selecting the data of the first and second areas so as not to exist. Arithmetic unit
In the measurement space, a plurality of nuclear magnetic resonance signal data arranged radially from the origin, including the origin, data of the first area within a predetermined range from the origin, and farther from the origin than the first area Filter setting means for selecting data in the second region including the range;
The nuclear magnetic resonance signal data of the first and second regions selected by the filter selection means and the nuclear magnetic resonance signal data of the entire measurement space are each subjected to inverse Fourier transform, and the first region inverse Fourier transform data and Inverse Fourier transform means for obtaining second region inverse Fourier transformed data and overall inverse Fourier transformed data,
Complex conjugate means for obtaining estimated data by obtaining complex conjugate data of the second region inverse Fourier transformed data;
Phase correction means for performing phase correction by subtracting or adding the phase value of the data after the first region inverse Fourier transform from the data after the second region inverse Fourier transform or the complex conjugate data; and
Adding means for obtaining a reconstructed image by adding the data after the whole inverse Fourier transform and the estimated data;
Image reconstruction program to function as

Images