CN101785684A - Ultrasonic endoscopic synthetic aperture imaging system and synthetic method of larger aperture - Google Patents

Abstract

The invention relates to an ultrasonic endoscopic synthetic aperture imaging system and a synthetic method of a larger aperture. The ultrasonic endoscopic synthetic aperture imaging system comprises a miniature ultrasonic probe, an ultrasonic probe drive circuit, a simulation receiving circuit, a digital processing circuit, a USB (Universal Serial Bus) interface circuit and a computer image display system. The synthetic method for obtaining the larger aperture is realized by adopting a coded excitation technology, a beam synthesis algorithm, a quadrature demodulation technology and a Cordic digital scanning conversion algorithm and concretely comprises the following steps of: encoded signal design and generation, beam synthesis (comprising longitudinal match filter, range migration correction and transverse match filter), quadrature demodulation and digital scanning conversion. Differing from a multi-array-element linear array and convex array scanning technology, the miniature ultrasonic probe of a single transducer array is adopted by the invention, is sent into a human body through an endoscopic biopsy forceps channel and is used for rotationally scanning in the body; meanwhile, the larger aperture is obtained through equivalent synthesis by utilizing the rotary characteristic of the probe, thereby greatly improving the signal-to-noise ratio and the transverse resolution of the system.

Description

The synthetic method of ultrasonic endoscopic synthesis aperture imaging system and larger aperture

[technical field]: the invention belongs to the ultrasonic endoscopic technical field of imaging, relate in particular to the beam synthesizing technology of ultrasonic receiving aperture.

[background technology]: ultrasonic endoscopic is sent the miniature ultrasonic scanheads into human body by the biopsy channel of fujinon electronic video endoscope, in endoscopic observation intracorporeal organ inner chamber mucosal surface, ultrasound probe obtains the faultage image of human viscera organ's wall, finding wherein early stage canceration and small tumor, is the best approach of current diagnosis human viscera organ pathological changes.Ultrasound wave has effect weakening when propagating in human body, along with the intensification of propagating the degree of depth, beam angle increases, and causes signal to noise ratio to reduce the lateral resolution variation.Because the beam angle and the angle of divergence and sound wave transmitting aperture are inversely proportional to, therefore, often adopt the synthetic aperture technique of many transducers array element, synthesize bigger aperture, sound wave interference stack by difference encourages transducer array element to make that each array element is sent constantly realizes the deflection and the focusing of emission acoustic beam; Again by to the echo signal of telecommunication of each transducer in addition dynamic deferred or phase compensation realize dynamic collectiong focusing, to obtain lateral resolution preferably.But ultrasonic endoscopic has strict requirement to the size of probe, and inflexibility length is less than 14mm, and diameter is less than 2.8mm, and the design of many transducers array element probe faces many difficulties, because size is bigger, is difficult to use in endoscope, is used for external supersonic more and detects.

[summary of the invention]: the objective of the invention is to solve the aforementioned problems in the prior, provide a kind of ultrasonic endoscopic synthesis aperture imaging system and larger aperture synthetic method that adopts single transducer array element, so that obtain and identical lateral resolution and the signal to noise ratio of many transducers of external supersonic array element scanning.

Ultrasonic endoscopic synthesis aperture imaging system based on single transducer array element provided by the invention comprises:

Miniature ultrasonic probe: the single transducer array element that adopts transceiver is used for receiving the ultrasonic reflection echo simultaneously as the ultrasonic emitting source, and is rotated scanning by miniature ultrasonic motor preposition driving transducer in human body.

Ultrasonic probe drive circuit: be connected with miniature ultrasonic probe by probe lead wire, adopt code-excited technology that the single transducer array element in the miniature ultrasonic probe is driven, improve transmitting power, strengthen backward energy.

Described ultrasonic probe drive circuit comprises code-excited circuit and match circuit, wherein code-excited circuit is made up of two groups of ultrasonic special chip MD1211 and TC6320, first group (U1, U2), MD1211 is according to the coded signal of FPGA output, the driving signal of output same-code, drive the code-excited signal of TC6320 output high-voltage square-wave, realize the power amplification of coded signal; Second group (U3, U4) is used for making last position of the coded signal of output to be stable at 0 current potential rapidly; Match circuit is made of series resistance R, shunt inductance L, and series resistance realizes the impedance matching between code-excited circuit and the transducer, and shunt inductance is realized the in parallel tuning of transducer; The high pressure coded signal of code-excited circuit output is after the match circuit coupling, and the excitation transducer sends the ultrasound wave with encoding characteristics.

Simulation receiving circuit: form by buffer circuit, amplifying circuit and filter circuit three parts, be connected with miniature ultrasonic probe by probe lead wire; Wherein, buffer circuit adopts the method for switching diode amplitude limit in parallel, utilizes its switching characteristic and in the operating characteristic in nonlinear operation district, exomonental amplitude is limited, and plays the protective effect to input amplifier; Amplifying circuit is finished two aspect tasks altogether, and the one, the faint ultrasonic signal that receives is amplified, be convenient to subsequent conditioning circuit and handle, the 2nd, ultrasonic signal is carried out gain compensation; Filter circuit adopts traditional RC circuit to finish the design of analog band-pass filter, realizes the removal and the anti-aliasing filter of high-frequency noise.Amplifying circuit wherein adopts the mode of two variable gain amplifier AD8331 cascade to realize the preposition amplification and the gain compensation of ultrasonic signal, preceding a slice adopts fixed 30dB gain, finish preposition enlarging function, back a slice adopts variable gain, by formula the signal amplitude curve of (1) is finished the gain compensation function, gain ranging 0～15.5dB:

S(t)＝R(x)I ₀e ^-2αx＝R(x)I ₀e ^-2αct (1)

Wherein, S (t) is an echo-signal, and R (x) is illustrated in the reflection coefficient at x place, and α represents mean attenuation coefficient, and c is the transmission speed of ultrasound wave in tissue, and to be ultrasound wave propagate into time of x, I from wave source to t ₀Be the incidence wave sound intensity.

Digital processing circuit: form by high-speed a/d change-over circuit, FPGA digital processing circuit and SRAM memory circuit, be connected with the simulation receiving circuit, finish ultrasonic signal digital processing and scan conversion function by data wire.

Usb circuit: be connected with digital processing circuit by data wire,

Computer picture display system: be used for realizing the demonstration of ultrasonoscopy and, realize the identification of computer, the reading of ultrasound data, the transmission of interface program operational order, and the communicating by letter of control USB device and computer to USB device to the operation of system.

During the work of ultrasonic endoscopic synthesis aperture imaging system, the miniature ultrasonic scanheads is sent into the human intracavity by the biopsy channel of fujinon electronic video endoscope, utilize the micromachine in the probe to drive single transducer array element rotation.In single transducer array element rotation, the high pressure coded pulse after the ultrasonic probe drive circuit output coupling encourages single transducer array element emission to have the ultrasonic signal of coding characteristic.Ultrasonic signal forms a ultrasonic echo sequence after the tissue reflection of different depth, receive and convert to the signal of telecommunication by former transducer array element.This signal of telecommunication receives, handles and form the Type B ultrasonoscopy by simulating receiving circuit and digital processing circuit, after USB interface input computer preservation and demonstration.

The synthetic method that the above-mentioned imaging system of employing provided by the invention obtains larger aperture comprises: design also generates coded signal, wave beam synthesizes (comprising vertical matched filtering, range migration correction and horizontal matched filtering), quadrature demodulation, four parts of digital scan conversion.

1st, design and generation coded signal, the ultrasound wave that transducer is sent has coding characteristic;

Coded signal of the present invention adopt 4 Barker (+1+1+1-1) sign indicating number is as the excitation coding, constitutes subpulse in the excitation coding with positive negative pulse stuffing, i.e. unit element, Baud Length are the 1/f reciprocal of the ultrasound wave mid frequency that sends of single transducer array element ₀, the first three groups positive negative pulse stuffing constitutes the front three of 4 Barker codes, and the negative positive pulse of last group constitutes last position of 4 Barker codes.

2nd, according to the coded signal characteristics of the 1st step design, design vertical matched filter, finish the longitudinal focusing of echo-signal in time domain and handle;

Described vertical matched filter is a series of characteristic parameter.Because ultrasound echo signal has coding characteristic, therefore, vertically the parameter of matched filter is corresponding with ultrasound echo signal, also has coding characteristic, and its parameter is the complex conjugate of ultrasound echo signal.Adopting vertical matched filter to focus on to be actually in time domain makes ultrasound echo signal and its complex conjugate do the process of convolution.In digital processing circuit, signal successively with the summation of multiplying each other of the parameter of vertical matched filter, τ, i.e. 1/f advance at every turn ₀, f ₀Be hyperacoustic mid frequency, obtain the result of matched filtering at last.

3rd, according to the rotation sweep characteristics of the miniature ultrasonic probe of single transducer array element, finish range migration correction, to remove laterally and 2D signal coupling longitudinally in time domain;

The method of described range migration correction is as follows:

On the polar coordinate image that collects, to arbitrary scanning line, it is the impact point of surveying that the point on it all can be regarded as.The distance of setting certain point and the center of rotation of single transducer array element on the scanning line is x, its value for this radius of turn of the abscissa value on the scanning line (surface of emission of representing single transducer array element over against this time distance) and single transducer array element with.The deflection of single transducer array element

After the angle, this distance of putting the surface of emission of single transducer array element is x ', and x, r, x ' become an Atria bar limit, by the cosine law, obtain:

R is the radius of turn of transducer, and x ' is deflection

After the angle, this abscissa value on scanning line, therefore, the range migration amount that range migration correction will compensate is:

Δx＝x′-x (3)

Set up search relationship with this, can obtain corresponding range migration amount according to anglec of rotation tolerance, realize range migration correction, it is last that the signal that will be positioned at different distance is adjusted to same distance.

4th, the signal edge behind the 3rd step range migration correction laterally is converted into frequency domain,, finishes transverse focusing in frequency domain and handle by horizontal matched filter;

Described horizontal matched filter is:

$h\left(t\right)=\exp \left\{\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HSA00000006906200012.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\right[f_{\mathrm{dc}}t-\frac{1}{2}{f}_{\mathrm{dr}}{t}^2\left]\right\},(-\frac{{\mathrm{<figure-callout\; id="2"\; label="T\; s"\; filenames="HSA00000006906200012.png"\; state="\{\{state\}\}">T</figure-callout>}}_s}{2}<t<\frac{T_s}{2})---\left(4\right)$

f _DcBe doppler centroid, f _DrBe doppler frequency rate.

5th, time domain is returned in the signal conversion after the transverse focusing processing of the 4th step, carry out quadrature demodulation, obtain reflecting the polar coordinate image of tissue feature, the abscissa of polar coordinate image is represented utmost point footpath, and vertical coordinate is represented polar angle;

The method of described quadrature demodulation is as follows:

Time domain is returned in signal conversion after above-mentioned the 4th step transverse focusing processing, and then ultrasonic signal is expressed as:

y(t)＝a(t)cos[ω ₀t+θ] (5)

Wherein, a (t) is the instantaneous amplitude of ultrasonic signal, ω ₀Be the ultrasonic carrier frequency, θ is the initial phase of signal; Ultrasonic signal converts the digital signal that is shown below to after the ADC sampling, represent that with a sequence n is a positive integer:

Wherein, T _s=1/f _s, f _sBe the ADC sample frequency, digital signal x (n) is ω with the frequency that digital controlled oscillator NCO exports respectively ₀Cosine and sine signal multiply each other, obtain the mutually orthogonal signal of two-way:

This two paths of signals is passed through wave digital lowpass filter, filtering 2 ω respectively ₀Frequency component, to obtain I, the orthogonal baseband signal of Q two-way:

The highest frequency of the baseband signal that filtering is obtained is subjected to the restriction of transducer bandwidth, generally lower, and sample rate still is the sample frequency of ADC, if directly carry out signal processing, the operand of circuit is very big, therefore need carry out data pick-up, reduce sample frequency, to reduce the operand of subsequent conditioning circuit; Two-way digital orthogonal baseband signal behind the data pick-up is:

Wherein, m asks its root-mean-square for the serial number after extracting, and the amplitude information that can obtain ultrasonic signal is a baseband signal:

$a\left(m\right)=\sqrt{I^2\left(m\right)+Q^2\left(m\right)}---\left(13\right).$

6th, adopt hardware Cordic algorithm to finish the digital scan conversion of polar coordinate image, so that eye-observation to the rectangular coordinate image.

Described as follows to the method for the digital scan conversion of rectangular coordinate image by polar coordinate image:

Adopt hardware Cordic algorithm to finish digital scan conversion, the Cordic algorithm is a kind of loop iteration algorithm, by continuous beat to a series of fixed angles relevant with the computing radix, the required position that rotates to of iterative approach, the vector of the angle that finally turns over and be polar angle in the polar coordinate; For with polar coordinate and rectangular coordinate mutual mapping, the object vector of regulation Cordic algorithm, promptly the final position of iterative approach is the X-axis of rectangular coordinate system, the iterative formula of utmost point footpath and polar angle is suc as formula (14), (15), (16),

x _i+1＝cos(θ _i)(x _i-y _i?tan(θ _i)) (14)

y _i+1＝cos(θ _i)(y _i-x _i?tan(θ _i)) (15)

$\mathrm{\&theta;}=\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}{\mathrm{\&theta;}}_i---\left(16\right)$

By beat iteration repeatedly, make object vector V ₁(x ₁, y ₁) rotate on the X-axis i.e. y in the iterative formula _I+1When being zero, iteration finishes, the x of this moment _I+1Be the utmost point and directly be worth, θ is the polar angle value;

Utilize the plus-minus of FPGA and shift operation to finish mathematical operation in the Cordic algorithm, at first produce in the image that need be presented under the rectangular coordinate system each pixel rectangular coordinate (X one to one by FPGA _i, Y _i), FPGA utilizes adder and shift register interative computation and the correction of finishing the Cordic algorithm of self, and (ρ θ), becomes each pixel address stored in SRAM with the polar coordinate information translation that obtains to obtain corresponding polar coordinate information under the control of FPGA; FPGA reads data among the SRAM according to this address, i.e. the gray value of respective pixel, and then form a width of cloth rectangular coordinate image.

Advantage of the present invention and good effect:

(1) multiple diagnostic.Both can carry out the ultrasonic scanning imaging again, obtain the histologic characteristics of each tomography of digestive organs tube wall by the pathological change form of fujinon electronic video endoscope direct observation mucomembranous surface.Therefore enlarge the diagnostic area of endoscope, improved the diagnosis capability of endoscope.

(2) long service life.Adopt the miniature ultrasonic motor preposition driving probe rotation of granted patent, need not easily broken soft syndeton, extend service life greatly.

(3) dynamic focusing.Owing to realized single probe synthetic aperture technique, solved the contradiction between investigation depth and the lateral resolution, make ultrasound wave lateral resolution on each degree of depth of propagating constant, can obtain diagnostic result more accurately.

(4) simple in structure.Endoscope compares with conventional ultrasound, and the preposition structure that has saved stone dead wire and external motor of motor has been simplified system; Adopt single transducer array element to be rotated scanning, probe size is little, in the time of the realization synthetic aperture, adopts the probe of bigger linear array or protruding battle array, only improves on imaging algorithm, makes the upgrading of system and maintenance flexible more, convenient.

[description of drawings]:

Fig. 1 is a ultrasonic endoscopic synthesis aperture imaging system theory diagram of the present invention.

Fig. 2 is the ultrasonic probe drive circuit that the present invention is based on code-excited technology.

Fig. 3 is a simulation receiving circuit of the present invention.

Fig. 4 is a ultrasonic digital processing system of the present invention.

Fig. 5 is a usb circuit of the present invention.

Fig. 6 is the transducer excitation coding that the present invention adopts.

Fig. 7 is that the single transducer array element rotation of the present invention receives sketch map.

Fig. 8 is a longitudinal focusing sketch map of the present invention.

Fig. 9 is a quadrature demodulation schematic diagram of the present invention.

Figure 10 is Cordic algorithm principle figure of the present invention.

Figure 11 is a digital scan conversion leak source interpolation sketch map of the present invention.

Among the figure, 1 is 90 ° of hard bendings, 2 biopsy channels, 3 medical electronic endoscopies, 4 rotors, 5 motor stators, 6 probe lead wires, 7 probing shells, 8 transducers, 9 ultrasonic beams, 10 single transducer element position I, 11 single transducer element position II.

[specific embodiment]:

Embodiment 1, ultrasonic endoscopic synthesis aperture imaging system

As shown in Figure 1, the ultrasonic endoscopic synthesis aperture imaging system based on single transducer array element provided by the invention comprises six parts: miniature ultrasonic probe (being made up of parts 4～8), ultrasonic probe drive circuit, simulation receiving circuit, digital processing circuit, usb circuit and computer picture display system.

Details are as follows for system's each several part:

1, miniature ultrasonic probe

The single transducer array element that adopts transceiver is as the ultrasonic emitting source, be used for receiving the ultrasonic reflection echo simultaneously, and (miniature ultrasonic probe has been authorized patent of invention, the patent No.: ZL200410019745.9) to be rotated scanning by miniature ultrasonic motor preposition driving transducer in human body.

2, ultrasonic probe drive circuit

Utilize the ultrasonic special chip MD1211 and the TC6320 of Supertex company, designed the ultrasonic probe drive circuit, as shown in Figure 2, comprise code-excited circuit and match circuit, wherein code-excited circuit is made up of two groups of ultrasonic special chip MD1211 and TC6320, as shown in Figure 2, first group (U1, U2), MD1211 export the driving signal of same-code according to the coded signal of FPGA output, drive the code-excited signal of TC6320 output high-voltage square-wave, realize the power amplification of coded signal; In the circuit, U1 converts the 3.3V coded signal of FPGA output to yard type anti-phase driving signal in this section.Drive P raceway groove in the U2 chip and N raceway groove by the anti-phase signal of this yard type then and take turns conducting, output high voltage level signal; Second group (U3, U4) is used for making last position of the coded signal of output to be stable at 0 current potential rapidly.In this part circuit, U3 is started working after finishing by U1, U2 conversion at coded signal, and this moment, U3 drove P or the conducting of N-channel MOS pipe among the U4, and stable output signal at 0V, is avoided the mistake excitation to single transducer array element.

When single transducer array element is worked, externally be rendered as capacitive near resonant frequency,, can cause the signal waveform distortion, reduce ultrasound emission efficient if directly link to each other with the code-excited signal of square wave.In addition, according to alternating-current circuit theory, when having only load to be the optimum load of driving source, just can obtain peak power output.But the electrical impedance of single transducer array element is different from the optimum load impedance of code-excited circuit, therefore must carry out impedance matching, so that transducer obtains peak power output.Match circuit be used for realizing to the tuning coupling of single transducer array element internal resistance and with the impedance matching of code-excited circuit, constitute by series resistance R, shunt inductance L, series resistance realizes the impedance matching between code-excited circuit and the transducer, and shunt inductance is realized the in parallel tuning of transducer.The high pressure coded signal of code-excited circuit output is after the match circuit coupling, and the excitation transducer sends the ultrasound wave with encoding characteristics.

3, simulation receiving circuit

The simulation receiving circuit is made up of three parts, and its structure comprises as shown in Figure 3: 1) buffer circuit.Because the ultrasonic endoscopic synthesis aperture imaging system adopts the ultrasonic transducer of transceiver, amplifying circuit needs and encoding transmitting circuit connects together, and therefore damages amplifying circuit for fear of the high pressure pumping signal, must increase buffer circuit before amplifying circuit.Adopt the method for switching diode amplitude limit in parallel, utilize its switching characteristic and, exomonental amplitude is limited in the 1V scope, play the protective effect of pair amplifier input in the operating characteristic in nonlinear operation district.2) amplifying circuit.Finish two aspect tasks altogether, the one, faint ultrasonic signal is amplified, be convenient to subsequent conditioning circuit and handle, the 2nd, ultrasonic signal is carried out gain compensation.When ultrasound wave was propagated in tissue, its energy was decayed according to index law along with the increase of propagation distance.Represent mean attenuation coefficient with α, be illustrated in the reflection coefficient at x place with R (x), then echo-signal S (t) can be expressed as

S(t)＝R(x)I ₀e ^-2αx＝R(x)I ₀e ^-2αct (1)

Wherein, c is the transmission speed of ultrasound wave in tissue, and t is ultrasound wave propagates into x from wave source time; I ₀Be the incidence wave sound intensity.The present invention adopts the scheme of two variable gain amplifier AD8331 cascade to realize the preposition amplification and the gain compensation of ultrasonic signal, preceding a slice adopts fixed 30dB gain, finish preposition enlarging function, back a slice adopts variable gain, signal amplitude curve by formula 1 is finished the gain compensation function, gain ranging 0～15.5dB.3) filter circuit.Adopt traditional RC circuit to finish the design of analog band-pass filter, system adopts the ultrasonic transducer of 8MHz, relative bandwidth 40%, so the free transmission range of wave filter is set at 6.4MHz～9.6MHz, the removal and the anti-aliasing filter of realization high-frequency noise.

4, digital processing circuit

Digital processing circuit is made up of high-speed a/d change-over circuit, FPGA digital processing circuit and SRAM memory circuit, finishes ultrasonic signal digital processing and scan conversion function.As shown in Figure 4.FPGA is the core of system, both has been responsible for sequencing contro, is responsible for Digital Signal Processing again.

5, usb circuit

As shown in Figure 5, usb circuit is realized by the ISP1581 interface chip, and is used enhancement mode 51 single-chip microcomputers as local cpu, carrying firmware program, the overall situation control of mouthful circuit that achieves a butt joint.Dma controller (DMAC) selects for use CPLD to realize, is responsible for sending and reading of DMA signal, and the usb data transmission under the control dma mode.Data transmission bauds is 8MB/s, and per second can show 5 width of cloth images.

6, computer picture display system

Image display system comprises interface program, driver, firmware downloads program.Interface program is used for realizing the demonstration of ultrasonoscopy and to the operation of system, as freezing of image, storage with print; Driver is realized the identification of computer to USB device, the reading of ultrasound data, and the transmission of interface program operational order; The firmware downloads program when USB interface is connected with computer automatically from downloaded to USB device, the communicating by letter of control USB device and computer.

The synthetic method of embodiment 2, acquisition larger aperture

The inventive method adopts single transducer array element to launch and receives ultrasound wave, utilizes the rotation effect of single transducer array element, and at the different moment and position emission and reception ultrasonic echo, equivalence " synthesizing " goes out bigger aperture.

Improve emitted energy with code-excited technology, solve single transducer array element transmitting power problem of smaller; Simultaneously in the synthetic aperture algorithm, when realizing transverse focusing, the coded signal of depth direction is also focused on processing.

This method specifically comprises: design also generates coded signal, and wave beam synthesizes (comprising vertical matched filtering, range migration correction and horizontal matched filtering), quadrature demodulation, four parts of digital scan conversion.Each several part is described below:

1, design and generation coded signal

Different with traditional pulse excitation, code-excited employing long codes pulse excitation transducer, the ultrasonic signal that excites is a long pulse.Because the persistent period of code-excited signal is longer than the pulse response time of transducer, thereby can increase the energy that ultrasonic signal carries.Code-excited reflection echo also is a long pulse, and this can reduce the longitudinal resolution of system, needs echo is focused on longitudinally, to obtain and the identical longitudinal resolution of pulse excitation, improves the signal to noise ratio of ultrasound echo signal simultaneously.

The present invention adopt 4 Barker (+1+1+1-1) sign indicating number is as excitation coding.In order to obtain higher emission effciency, the frequency spectrum of coded signal is dropped in the bandwidth range of ultrasonic transducer as much as possible.Therefore, the present invention modulates the subpulse of boot code, constitute the unit element of coded signal with positive negative pulse stuffing, Baud Length is the 1/f0 reciprocal of transducer mid frequency, the frequency spectrum of coded signal is overlapped with the frequency spectrum of single transducer array element to greatest extent, and to obtain maximum emission effciency, the coded signal after the modulation as shown in Figure 6, the first three groups positive negative pulse stuffing constitutes the front three of 4 Barker codes, and the negative positive pulse of last group constitutes last position of 4 Barker codes.

2, wave beam is synthetic

Wave beam is synthetic to be the core of aperture synthetic method.Adopt the principle of single transducer array element probe synthetic aperture as follows: for a single point target, to the echo of reflection disperse thereafter, therefore transducer array element can receive the echo of this point target reflection in certain anglec of rotation in the rotation sweep process.According to this characteristics of rotation sweep, in the anglec of rotation that can receive the point target echo, this revolving property equivalence can be become bigger aperture, its theoretical lateral resolution is D/2, wherein, D is the transmitting aperture of transducer.

The synthetic concrete steps of wave beam are:

1) echo-signal with encoding characteristics is carried out longitudinal focusing.The present invention adopts Barker code excitation transducer, and therefore, vertically the focusing of matched filter is actually the process of a time domain convolution, as shown in Figure 7.Signal successively with the summation of multiplying each other of the parameter of matched filter, the τ that at every turn advances, i.e. 1/f, f is hyperacoustic mid frequency, obtains the result of matched filtering at last.Coded signal the same result in the time of after longitudinal focusing, can obtaining encouraging with pulse, but its amplitude and signal to noise ratio will be significantly better than the latter, and its signal to noise ratio has promoted 1.36dB at least.

2) time domain range migration correction.Transducer is when the echo of acceptance point target, because the difference of position, makes transducer different with distance between point target, and therefore, the echo that receives has different time delays and phase place to change, as shown in Figure 8.If do not handle, the image that point target obtains will be one and be parabolic disc of confusion that this is an amount with two dimensional character, when transverse focusing compresses, can vertically produce component, thereby influence longitudinal resolution, therefore, parabola will be proofreaied and correct before transverse compression is straight line.According to single probe rotation characteristics, in Fig. 8, R be single transducer array element during over against sensing point sensing point to the distance of transducer, R ' is the distance after rotating to an angle, r is the radius of turn of transducer, R ', R+r, r have constituted leg-of-mutton three limits, obtain R ' and just can obtain the range curvature amount of sensing point in time domain.

The specific implementation method is that on the polar coordinate image that collects, to arbitrary scanning line, it is the impact point of surveying that the point on it all can be regarded as.The distance of setting certain point and the center of rotation of single transducer array element on the scanning line is x, its value for this radius of turn of the abscissa value on the scanning line (surface of emission of representing single transducer array element over against this time distance) and single transducer array element with.The deflection of single transducer array element

After the angle, this distance of putting the surface of emission of single transducer array element is x ', and x, r, x ' become an Atria bar limit, by the cosine law, obtain:

R is the radius of turn of transducer, and x ' is deflection After the angle, this abscissa value on scanning line, therefore, the range migration amount that range migration correction will compensate is:

Δx＝x′-x (3)

Set up search relationship with this, can obtain corresponding range migration amount according to anglec of rotation tolerance, realize range migration correction, it is last that the signal that will be positioned at different distance is adjusted to same distance.

3) signal of migration after proofreading and correct of adjusting the distance carries out horizontal focusing.The matched filter of transverse focusing should be identical with the character of horizontal ultrasonic echo, because transducer is a received signal in rotary course, therefore, horizontal echo-signal has Doppler effect, can be approximately linear FM signal.Therefore, it is as follows to design horizontal matched filter:

$h\left(t\right)=\exp \left\{\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HSA00000006906200012.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\right[f_{\mathrm{dc}}t-\frac{1}{2}{f}_{\mathrm{dr}}{t}^2\left]\right\},(-\frac{T_s}{2}<t<\frac{T_s}{2})---\left(4\right)$

f _DcBe doppler centroid, f _DrBe doppler frequency rate.The index of h (t) is the quadratic function of t, and having represented horizontal signal is a linear FM signal, uses it to describe the feature of horizontal signal.Because laterally the parameter of matched filter is more, it is bigger to handle operand in time domain, is difficult to guarantee real-time, therefore, at first the signal edge is laterally transformed to frequency domain, carry out horizontal matched filtering then, be transformed to time domain at last again, obtain the synthetic end product of wave beam.

3, quadrature demodulation

The ultrasonic endoscopic synthesis aperture imaging system adopts the B-mode ultrasonography pattern, utilizes the brightness of amplitude (envelope) the modulation image demonstration of ultrasonic signal, and the present invention adopts digital down-conversion technology to realize the extraction of signal amplitude information, and its principle as shown in Figure 9.System adjusts to the A/D change-over circuit front of quadrature demodulation, digitized in Mid Frequency realization echo-signal utilizes digital technology signal to be carried out processing such as mixing, filtering, sampling rate conversion then, to realize removing carrier wave, extract the purpose of baseband signal, realize the Digital Down Convert of signal.

Ultrasonic signal after wave beam synthesizes among Fig. 9 is expressed as:

y(t)＝a(t)cos[ω ₀t+θ] (5)

Wherein, a (t) is the instantaneous amplitude of ultrasonic signal, ω ₀Be the ultrasonic carrier frequency, θ is the initial phase of signal.Ultrasonic signal converts the digital signal that is shown below to after the ADC sampling:

Wherein, T _s=1/f _s, f _sBe the ADC sample frequency.Digital signal x (n) is ω with the frequency that digital controlled oscillator NCO exports respectively ₀Cosine and sine signal multiply each other, obtain the mutually orthogonal signal of two-way:

This two paths of signals is passed through wave digital lowpass filter, filtering 2 ω respectively ₀Frequency component, to obtain I, the orthogonal baseband signal of Q two-way:

The highest frequency of the baseband signal that filtering is obtained is subjected to the restriction of transducer bandwidth, generally lower, and sample rate still is the sample frequency of ADC, if directly carry out signal processing, the operand of circuit is very big, therefore need carry out data pick-up, reduce sample frequency, to reduce the operand of subsequent conditioning circuit.Two-way digital orthogonal baseband signal behind the data pick-up is:

Ask its root-mean-square, can obtain the amplitude information (baseband signal) of ultrasonic signal:

$a\left(m\right)=\sqrt{I^2\left(m\right)+Q^2\left(m\right)}---\left(13\right).$

4, digital scan conversion

The ultrasonic endoscopic synthesis aperture imaging system adopts the Type B imaging pattern, and promptly ultrasonic probe obtains the tomography gray level image of target object in the mode of sector scanning.Under this scan mode, the echo-signal of all angles is unfavorable for eye-observation with the polar form storage, therefore, must carry out the coordinate scan conversion to scanogram, and its form by rectangular coordinate is shown.In order to satisfy the requirement that image shows in real time, the present invention adopts hardware Cordic algorithm to finish digital scan conversion.

The Cordic algorithm is a kind of loop iteration algorithm, by continuous beat to a series of fixed angles relevant with the computing radix, and the required position that rotates to of iterative approach, the vector of the angle that finally turns over and be polar angle in the polar coordinate; For with polar coordinate and rectangular coordinate mutual mapping, the object vector of regulation Cordic algorithm, promptly the final position of iterative approach is the X-axis of rectangular coordinate system, the iterative formula of utmost point footpath and polar angle is suc as formula (14), (15), (16),

x _i+1＝cos(θ _i)(x _i-y _i?tan(θ _i)) (14)

y _i+1＝cos(θ _i)(y _i-x _i?tan(θ _i)) (15)

$\mathrm{\&theta;}=\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}{\mathrm{\&theta;}}_i---\left(16\right)$

By beat iteration repeatedly, make original vector V ₁(x ₁, y ₁) rotate on the X-axis i.e. y in the iterative formula _I+1When being zero, iteration finishes, the x of this moment _I+1Be the utmost point and directly be worth, θ is the polar angle value.The present invention utilizes the plus-minus of FPGA and shift operation to finish complex mathematical computings such as multiplication in the Cordic algorithm, trigonometric function, and utilizes the bin of FPGA to realize pile line operation.Thereby in hardware, finish digital scan conversion.At first produce in the image to be presented under the rectangular coordinate system each pixel rectangular coordinate (X one to one by FPGA _i, Y _i), FPGA utilizes adder and shift register interative computation and the correction of finishing the Cordic algorithm of self, and (ρ θ), becomes each pixel address stored in SRAM with the polar coordinate information translation that obtains to obtain corresponding polar coordinate information under the control of FPGA.FPGA reads data among the SRAM according to this address, i.e. the gray value of respective pixel, and then form a width of cloth rectangular coordinate image.

The rectangular coordinate image that obtains through scan conversion in the sampled data of radial direction than comparatively dense, but along there being big distance between the sampled data on the angle direction, especially in the far field.Therefore, need the image after the scan conversion is carried out the leak source interpolation.Algorithm of the present invention only carries out the interpolation of far field angle direction, i.e. the interpolation of far field circle.As shown in figure 11, and P (i, j), (i, j+1) two points for having same radius on the adjacent scanning lines are obtained the interpolation data of some points between them to P on the circumference at their places by the linear interpolation principle.Four the most contiguous some P (i, j), P (i, j+1), P (i+1, j), (i+1, j+1) sector of Zu Chenging is divided into four uniform little sector a, b, c, d to P.Which little sector display pixel drops in, just get the upper left interpolation data value in this little sector as its gray-scale displayed value, this method has been avoided only carrying out the angle direction distortion that causes under the one dimension interpolation situation radially and has been carried out the complex calculation that two-dimentional interpolation needs, and but can reach the effect that is similar to two-dimentional interpolation.

Concrete application example 1

Miniature ultrasonic probe enters human body through the biopsy forceps road of fujinon electronic video endoscope, esophagus is rotated scanning detects, and single transducer array element is fixed on the interior ultrasound electric machine of probe, and wherein single transducer array element is of a size of 2mm * 4mm.Esophagus is detected, when obtaining its surface image, also can obtain its interior tissue faultage image.By the aperture that the rotation effect of single transducer array element probe " is synthesized ", make the signal to noise ratio of system imaging promote 5.65dB, lateral resolution is 2.2mm.Wherein, the lifting of signal to noise ratio is divided into two parts, and vertically matched filtering has promoted 1.36dB, and laterally matched filtering has promoted 4.29dB; Point target can reach 15mm in the disc of confusion arc length that the far field of image forms, and through synthetic aperture processing, it can be compressed to 2.2mm, has significantly improved lateral resolution.

Concrete application example 2

Coat of the stomach is rotated scanning.Can realize complete detection by the joystick control ultrasonic probe position and the deflection angle under one's belt of fujinon electronic video endoscope to coat of the stomach.Because single probe synthetic aperture technique has realized hyperacoustic dynamic focusing, even the coat of the stomach of the some directions of ultrasonic probe distance is far away, also can obtain identical lateral resolution, can reach 2.2mm, and, needn't worry the distortion of image along with the increase lateral resolution of investigation depth is constant.

Concrete application example 3

Be illustrated in figure 1 as ultrasonic Synthetic Aperture System structure, can dynamically change the parameter of synthetic aperture algorithm wherein according to the requirement of the complexity that the real-time performance of system imaging and processing, the detection of a target are handled, comprise the matched filter parameter of revising depth direction and horizontal direction according to code length, rotation sweep speed, to realize the unification of focusing performance.The invention provides the data caching function, can in freeze frame, keep the synthetic preceding initial data in aperture, be convenient to the image of fixed position is carried out more accurate synthetic aperture processing.

Concrete application example 4

System structure is identical with application example 3, for increasing hyperacoustic penetration depth, can improve the length of code-excited signal under the prerequisite that the time domain waveform aliasing do not occur, or the sign indicating number type of change boot code, strengthens the transmitting power and the efficient of transducer.The average thickness of human body coat of the stomach is about 3.8mm, the boundary layer that has 5 character to differ from one another, the average thickness of each layer is about 0.76mm, therefore, under the situation that adopts the single transducer array element probe of 8MHz, code length is the longest to be 8, owing to the special figure place requirement that has of Barker code, therefore, add long codes and generally get 5 and 7; Along with the raising of frequency probe, code length can suitably extend.Simultaneously, change the depth direction matched filter parameter of synthetic aperture algorithm, comprise number of parameters and parameter value, identical with code length and encoding phase, to adapt to adjustment to pumping signal.So both increase investigation depth, do not influenced the resolution of system again.

Claims (10)

1. ultrasonic endoscopic synthesis aperture imaging system based on single transducer array element is characterized in that this system comprises:

Miniature ultrasonic probe: the single transducer array element that adopts transceiver is used for receiving the ultrasonic reflection echo simultaneously as the ultrasonic emitting source, and is rotated scanning by miniature ultrasonic motor preposition driving transducer in human body;

Ultrasonic probe drive circuit: be connected with miniature ultrasonic probe by probe lead wire, adopt code-excited technology that the single transducer array element in the miniature ultrasonic probe is driven, improve transmitting power, strengthen backward energy;

Simulation receiving circuit: form by buffer circuit, amplifying circuit and filter circuit three parts successively, be connected with miniature ultrasonic probe by probe lead wire; Wherein, buffer circuit adopts the method for switching diode amplitude limit in parallel, utilizes its switching characteristic and in the operating characteristic in nonlinear operation district, exomonental amplitude is limited, and plays the protective effect to input amplifier; Amplifying circuit is finished two aspect tasks altogether, and the one, the faint ultrasonic signal that receives is amplified, be convenient to subsequent conditioning circuit and handle, the 2nd, ultrasonic signal is carried out gain compensation; Filter circuit adopts traditional RC circuit to finish the design of analog band-pass filter, realizes the removal and the anti-aliasing filter of high-frequency noise;

Digital processing circuit: form by high-speed a/d change-over circuit, FPGA digital processing circuit and SRAM memory circuit, be connected with the simulation receiving circuit, finish ultrasonic signal digital processing and scan conversion function by data wire;

Usb circuit: be connected with the computer picture display system with digital processing circuit respectively by data wire, be used for transmit image data;

Computer picture display system: be used for realizing the real-time demonstration of ultrasonoscopy.

2. imaging system according to claim 1, it is characterized in that described ultrasonic probe drive circuit comprises code-excited circuit and match circuit, wherein code-excited circuit is made up of two groups of ultrasonic special chip MD1211 and TC6320, first group (U1, U2), MD1211 is according to the coded signal of FPGA output, the driving signal of output same-code drives the code-excited signal of TC6320 output high-voltage square-wave, realizes the power amplification of coded signal; Second group (U3, U4) is used for making last position of the coded signal of output to be stable at 0 current potential rapidly; Match circuit is made of series resistance R, shunt inductance L, and series resistance realizes the impedance matching between code-excited circuit and the transducer, and shunt inductance is realized the in parallel tuning of transducer; The high pressure coded signal of code-excited circuit output is after the match circuit coupling, and the excitation transducer sends the ultrasound wave with encoding characteristics.

3. imaging system according to claim 1, it is characterized in that the amplifying circuit in the described simulation receiving circuit adopts the mode of two variable gain amplifier AD8331 cascade to realize the preposition amplification and the gain compensation of ultrasonic signal, preceding a slice adopts fixed 30dB gain, finish preposition enlarging function, back a slice adopts variable gain, by formula the signal amplitude curve of (1) is finished the gain compensation function, gain ranging 0～15.5dB:

S(t)＝R(x)I ₀e ^-2αx＝R(x)I ₀e ^-2αct (1)

Wherein, S (t) is an echo-signal, and R (x) is illustrated in the reflection coefficient at x place, and α represents mean attenuation coefficient, and c is the transmission speed of ultrasound wave in tissue, and to be ultrasound wave propagate into time of x, I from wave source to t ₀Be the incidence wave sound intensity.

4. synthetic method that adopts the described imaging system of claim 1 to obtain larger aperture is characterized in that this method comprises:

1st, design and generation coded signal, the ultrasound wave that transducer is sent has coding characteristic;

2nd, according to the coded signal characteristics of the 1st step design, design vertical matched filter, finish the longitudinal focusing of echo-signal in time domain and handle;

3rd, according to the rotation sweep characteristics of the miniature ultrasonic probe of single transducer array element, finish range migration correction, to remove laterally and 2D signal coupling longitudinally in time domain;

4th, the signal edge behind the 3rd step range migration correction laterally is converted into frequency domain,, finishes transverse focusing in frequency domain and handle by horizontal matched filter;

5th, time domain is returned in the signal conversion after the transverse focusing processing of the 4th step, carry out quadrature demodulation, obtain reflecting the polar coordinate image of tissue feature, the abscissa of polar coordinate image is represented utmost point footpath, and vertical coordinate is represented polar angle;

6th, adopt hardware Cordic algorithm to finish the digital scan conversion of polar coordinate image, so that eye-observation to the rectangular coordinate image.

5. method according to claim 4, it is characterized in that described coded signal of the 1st step adopts 4 Barker codes as the excitation coding, constitute subpulse in the excitation coding with positive negative pulse stuffing, i.e. unit element, Baud Length are the 1/f reciprocal of the ultrasound wave mid frequency that sends of single transducer array element ₀, the first three groups positive negative pulse stuffing constitutes the front three of 4 Barker codes, and the negative positive pulse of last group constitutes last position of 4 Barker codes.

6. method according to claim 4 is characterized in that described vertical matched filter of the 2nd step is a series of characteristic parameter, and corresponding with ultrasound echo signal, its parameter is the complex conjugate of ultrasound echo signal; Adopting vertical matched filter to focus on to be actually in time domain makes ultrasound echo signal and its complex conjugate do the process of convolution; In digital processing circuit, signal successively with the summation of multiplying each other of the parameter of vertical matched filter, τ, i.e. 1/f advance at every turn ₀, f ₀Be hyperacoustic mid frequency, obtain the result of matched filtering at last.

7. method according to claim 4 is characterized in that the method for described range migration correction of the 3rd step is as follows:

On the polar coordinate image that collects, to arbitrary scanning line, it is the impact point of surveying that the point on it all can be regarded as; The distance of setting certain point and the center of rotation of single transducer array element on the scanning line is x, its value for the radius of turn of this abscissa value and single transducer array element on scanning line with, the abscissa value distance of the surface of emission during of representing single transducer array element wherein over against this; The deflection of single transducer array element

After the angle, this distance of putting the surface of emission of single transducer array element is x ', and x, r, x ' become an Atria bar limit, by the cosine law, obtain:

R is the radius of turn of transducer, and x ' is deflection

After the angle, this abscissa value on scanning line, therefore, the range migration amount that range migration correction will compensate is:

Δx＝x′-x (3)

Set up search relationship with this, can obtain corresponding range migration amount according to anglec of rotation tolerance, realize range migration correction, it is last that the signal that will be positioned at different distance is adjusted to same distance.

8. method according to claim 4 is characterized in that described horizontal matched filter of the 4th step is:

$h\left(t\right)=\exp \left\{j2\mathrm{\&pi;}\right[f_{\mathrm{dc}}t-\frac{1}{2}{f}_{\mathrm{dr}}{t}^2\left]\right\}(-\frac{T_s}{2}<t<\frac{T_s}{2})---\left(4\right)$

f _DcBe doppler centroid, f _DrBe doppler frequency rate.

9. method according to claim 4 is characterized in that the method for the 5th described quadrature demodulation of step is as follows:

Time domain is returned in signal conversion after the 4th step of claim 4, transverse focusing was handled, and then ultrasonic signal is expressed as:

y(t)＝a(t)cos[ω ₀t+θ] (5)

Wherein, a (t) is the instantaneous amplitude of ultrasonic signal, ω ₀Be the ultrasonic carrier frequency, θ is the initial phase of signal; Ultrasonic signal converts the digital signal that is shown below to after the ADC sampling, represent that with a sequence n is a positive integer:

Wherein, T _s=1/f _s, f _sBe the ADC sample frequency, digital signal x (n) is ω with the frequency that digital controlled oscillator NCO exports respectively ₀Cosine and sine signal multiply each other, obtain the mutually orthogonal signal of two-way:

This two paths of signals is passed through wave digital lowpass filter, filtering 2 ω respectively ₀Frequency component, to obtain I, the orthogonal baseband signal of Q two-way:

The highest frequency of the baseband signal that filtering is obtained is subjected to the restriction of transducer bandwidth, generally lower, and sample rate still is the sample frequency of ADC, if directly carry out signal processing, the operand of circuit is very big, therefore need carry out data pick-up, reduce sample frequency, to reduce the operand of subsequent conditioning circuit; Two-way digital orthogonal baseband signal behind the data pick-up is:

Wherein, m asks its root-mean-square for the serial number after extracting, and the amplitude information that can obtain ultrasonic signal is a baseband signal:

$a\left(m\right)=\sqrt{I^2\left(m\right)+Q^2\left(m\right)}---\left(13\right).$

10. method according to claim 4 is characterized in that the 6th step is described as follows to the method for the digital scan conversion of rectangular coordinate image by polar coordinate image:

Adopt hardware Cordic algorithm to finish digital scan conversion, the Cordic algorithm is a kind of loop iteration algorithm, by continuous beat to a series of fixed angles relevant with the computing radix, the required position that rotates to of iterative approach, the vector of the angle that finally turns over and be polar angle in the polar coordinate; For with polar coordinate and rectangular coordinate mutual mapping, the object vector of regulation Cordic algorithm, promptly the final position of iterative approach is the X-axis of rectangular coordinate system, the iterative formula of utmost point footpath and polar angle is suc as formula (14), (15), (16),

x _i+1＝cos(θ _i)(x _i-y _i?tan(θ _i)) (14)

y _i+1＝cos(θ _i)(y _i-x _i?tan(θ _i)) (15)

$\mathrm{\&theta;}=\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}{\mathrm{\&theta;}}_i---\left(16\right)$

By beat iteration repeatedly, make object vector V ₁(x ₁, y ₁) rotate on the X-axis i.e. y in the iterative formula _I+1When being zero, iteration finishes, the x of this moment _I+1Be the utmost point and directly be worth, θ is the polar angle value;

Utilize the plus-minus of FPGA and shift operation to finish mathematical operation in the Cordic algorithm, at first produce in the image that need be presented under the rectangular coordinate system each pixel rectangular coordinate (X one to one by FPGA _i, Y _i), FPGA utilizes adder and shift register interative computation and the correction of finishing the Cordic algorithm of self, and (ρ θ), becomes each pixel address stored in SRAM with the polar coordinate information translation that obtains to obtain corresponding polar coordinate information under the control of FPGA; FPGA reads data among the SRAM according to this address, i.e. the gray value of respective pixel, and then form a width of cloth rectangular coordinate image.

Images