CN103536316B - Method for self-adaptation ultrasonic imaging of spatio-temporally smoothed coherence factor type - Google Patents

Abstract

The invention discloses a method for self-adaptation ultrasonic imaging of a spatio-temporally smoothed coherence factor type. The method comprises the following steps that (1) beam forming preprocessing is conducted; (2) beam forming processing is conducted, wherein beam forming processing comprises the steps that delayed focusing is conducted on each channel received signal, a coherence factor value of the spatio-temporally smoothed coherence factor type is calculated according to the channel received signal after delayed processing, and weighted processing is conducted on the coherence sum of the received signals through coherence factors so that beams of each scanning line can be obtained and output; (3) beam forming post-processing is conducted, wherein beam forming post-processing comprises the steps that envelope detection, logarithm compression, scanning conversion and display are conducted on the scanning lines. According to the method for self-adaptation ultrasonic imaging of the spatio-temporally smoothed coherence factor type, side lobe can be restrained, sound disorder is reduced, the contrast ratio of an image is improved, an artifact is remarkably eliminated, the imaging quality is improved, the calculation complexity is low, and the method is easy to achieve in hardware.

Description

A kind of level and smooth coherence factor class adaptive ultrasound imaging method during sky

Technical field

The present invention relates to ultrasonic imaging technique, level and smooth coherence factor class adaptive ultrasound imaging method when being specifically related to a kind of empty.

Background technology

Ultra sonic imaging, as the important component part of modern medicine image technology, is widely used in clinical diagnosis with treatment.Compared with other imaging techniques, it is advantageous that noinvasive, without ionizing radiation, easy to use, real-time and low price etc.But ultra sonic imaging quality, as spatial resolution, contrast resolution, frame per second etc., generally unsatisfactory, be further improved.Wave beam forming is in core position in whole ultrasonic image-forming system, plays decisive role to image quality.At present, what be widely adopted in ultrasonic image-forming system is time delay superposition (delay-and-sum, DAS) beam-forming technology.Although it simply and effectively can realize the reconstruction of ultrasonoscopy, but its wave beam formed has wider main lobe width and higher side lobe levels, result in lower spatial resolution and more weak secondary lobe interference rejection capability.Traditional Apodization techniques uses one group of predetermined weight (as Hamming window) to reduce secondary lobe interference, but can widen main lobe, namely sacrifices certain spatial resolution.

In recent years, Adaptive beamformer technology had become the study hotspot in ultra sonic imaging field, was used to reduce side lobe levels simultaneously and reduce main lobe width." self adaptation " refers to that its weight is determined by the relevant information extracted from Received signal strength.One class methods determine weighing vector according to some optimization criterions.As minimum variance (minimum variance, MV) beam-forming technology.These class methods can improve spatial resolution and contrast preferably, but have higher computation complexity, and its hardware implementing is more difficult.Another kind of method is by analyzing main lobe signal and building a weighter factor from the feature of axis signal, thus suppresses to strengthen main lobe signal from axis signal.Coherence factor (the coherence factor designed as by the coherence measured between Received signal strength, CF), and based on the coherence of the phase information of Received signal strength and the phase coherence factor (PCF) built and symbol coherence factor (SCF) etc.CF be defined as the relevant of Received signal strength in aperture and energy and the ratio of total irrelevant energy.Its definition is as follows:

$\mathrm{CF}\left[n\right]=\frac{{\left|{\mathrm{\Σ}}_{m=0}^{M-1}{x}_m\right[n\left]\right|}^2}{M{\mathrm{\Σ}}_{m=0}^{M-1}{\left|x_m\right[n\left]\right|}^2},$

Wherein, M receives array element (passage) sum, x _m[n] is that n is the time index (time index) of signal, and its value is between [0,1] at the Received signal strength applying m the passage focused on after time delay.After focusing, highly will be concerned with from signal on the axle of scattering point on main lobe, obtain high CF value, and be incoherent from secondary lobe target from axis signal, produce low CF value.Like this, the effect of suppressed sidelobes can be played by CF sum weighting to received signal.The full name of SCF is sign coherence factor, can regard a kind of special case of PCF as.It designs based on the coherence messages of the sign bit of Received signal strength, is defined as:

SCF ^p[n]=|1-σ| ^p，and

$\mathrm{\σ}=\sqrt{1-{\left[\frac{1}{M}{\mathrm{\Σ}}_{m=0}^{M-1}{b}_m\left[n\right]\right]}^2},$

Wherein b _m[n] is Received signal strength x after time delay _mthe sign bit of [n], namely works as x _mduring [n]>=0, b _m[n] is 1, works as x _mbe-1 during [n] < 0, index p is adjustable parameter.Similarly, the dispersion (standard deviation) from the phase place (or symbol) of signal on the axle of main lobe target is 0, produces high factor values, and increasing from axis signal phase dispersion degree from secondary lobe target, cause low factor values.This kind of method based on coherence measurement can suppressed sidelobes preferably, and generally there is lower amount of calculation and can in imaging system real-time implementation.

Unfortunately, the image that coherence factor class formation method obtains can produce artifact (artifacts), particularly when low signal-to-noise ratio and low Signal to Interference plus Noise Ratio.Such as, can produce black cavity around strong scatterer, the speckle variance (specklevariance) in homogeneous background region can increase, and integral image brightness reduces even texure information dropout etc.Artifact destroys the spot pattern (speckle pattern) of background area, also can reduce picture quality simultaneously.Therefore, under ensureing that this kind of adaptive weighted factor suppressed sidelobes improves the effect of contrast, how to strengthen them to the vigorousness of low signal-to-noise ratio and eliminate artifact, become technical problem urgently to be resolved hurrily.

Summary of the invention

For the defect of prior art, the object of the present invention is to provide a kind of empty time level and smooth coherence factor class adaptive ultrasound imaging method.

According to the present invention, level and smooth coherence factor class adaptive ultrasound imaging method when providing a kind of empty, comprises the following steps:

(1) Wave beam forming pre-treatment;

(2) Wave beam forming process, comprise and delay and focusing is carried out to each channel receiving signal, the coherence factor value of level and smooth coherence factor class when calculating empty according to the channel receiving signal after delay process, utilize this coherence factor to received signal relevant and do weighting process, the Wave beam forming obtaining every bar scanning line exports;

(3) Wave beam forming post processing, comprises and carries out envelope detected, logarithmic compression, scan conversion and display to scanning line.

Compared with prior art, the present invention has following advantage:

1, remain the suppressed sidelobes of coherence factor class self adaptive imaging method (as CF, SCF) and the advantage of minimizing sound mixed and disorderly (clutter), improve the contrast of image;

2, can artifact be eliminated significantly and improve image quality;

3, computation complexity is lower, is easy to realize within hardware.

Accompanying drawing explanation

With reference to explanation below, by reference to the accompanying drawings, best understanding can be had to the present invention.In the accompanying drawings, identical part can be represented by identical label.

Fig. 1 be application the present invention carry empty time level and smooth coherence factor a typical phased array imaging flow chart;

Fig. 2 is the cyst body mould emulating image utilizing different beams formation technology to obtain;

Fig. 3 is the emulating image obtained based on the corrective measure of StS-CF proposed by the invention;

Fig. 4 is the image of the geabr experimental data utilizing different beams formation technology to obtain.

Detailed description of the invention

In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with accompanying drawing and exemplary embodiment, the present invention is further elaborated.Should be appreciated that exemplary embodiment described herein only in order to explain the present invention, the scope of application be not intended to limit the present invention.

Figure 1 shows that based on proposed by the invention empty time level and smooth coherence factor class Adaptive beamformer technology a typical ultra sonic imaging flow chart.In this particular example, phased array imaging pattern is employed.But should be appreciated that the present invention is also applicable to other imaging patterns, such as synthetic aperture is ultrasonic.Without loss of generality, this ultrasonic imaging method comprises the following steps:

(1) Wave beam forming pre-treatment, comprises setting sonac transmitting and receiving pattern, carries out digitized, amplification, filtering etc. to received signal;

(2) Wave beam forming process, comprise and delay and focusing (generally carrying out Dynamic receive focus) is carried out to each channel receiving signal, the coherence factor value of level and smooth coherence factor class when calculating this sky according to the channel signal after delay process, channel signal relevant and the Wave beam forming obtaining every bar scanning line after coherence factor weighting process export;

(3) Wave beam forming post processing, comprises and carries out the processes such as envelope detected, logarithmic compression, scan conversion and display to scanning line.

Below, first the image-forming principle of level and smooth coherence factor during empty in the present invention involved by wave beam formation processing is analyzed and explained.

The present invention propose empty time level and smooth coherence factor shown in (spatio-temporally smoothedcoherence factor, StS-CF) be defined as follows:

$\mathrm{StS}-\mathrm{CF}\left[n\right]=\frac{{\mathrm{\&Sigma;}}_{k=-K}^K{\left|{\mathrm{\&Sigma;}}_{l=0}^{M-L}{\mathrm{\&Sigma;}}_{m=l}^{L+l-1}{x}_m\right[n+k\left]\right|}^2}{(M-L+1){\mathrm{\&Sigma;}}_{k=-K}^K{\mathrm{\&Sigma;}}_{l=0}^{M-L}{\left|{\mathrm{\&Sigma;}}_{m=l}^{L+l-1}{x}_m\right[n+k\left]\right|}^2},$

Wherein, M is the sum of receiving sensor array element, x _m[n+k] is the signal (discrete form) after time delay that m array element (passage) receives, n and n+k in square brackets is time index sequence number, L and K is two argument of type integers.

Different from existing CF, StS-CF contains space smoothing (spatial smoothing) process, and be divided into the subarray of M-L+1 overlap by receiving array, each submatrix shows L array element.Meanwhile, it also contains time average (time averaging) process, namely measures the signal coherence degree of a pulse persistance (containing 2*K+1 time samples).Because rear scatter echo can be considered to originate from same transmitted pulse signal, therefore, typically, 2*K+1 is taken as the length of transmitted pulse signal.In a word, the definition of StS-CF be based on several sub-array beam and between the coherence measurement of one section of pulse signal, by above-mentioned empty time smoothing process be incorporated in CF.

Parameter L value, between 1 and M, can be used as user-defined parameter to adjust imaging performance and the vigorousness of the method.When not considering time average (i.e. K=0), when L gets 1, StS-CF is transformed into CF, and StS-CF perseverance is 1 when L gets M, is equivalent to not to relevant and do weighting process.L can apply according to the ultra sonic imaging of reality (as the detected object paid close attention to) select to determine.Preferably, the desirable M/2 of L.

Being concerned with this factor pair Received signal strength and doing weighting process, namely obtain the output of Wave beam forming:

$y\left[n\right]=\mathrm{StS}-\mathrm{CF}\left[n\right]*y_{\mathrm{das}}\left[n\right]=\mathrm{StS}-\mathrm{CF}\left[n\right]*\frac{1}{M}{\mathrm{\&Sigma;}}_{m=0}^{M-1}{x}_m\left[n\right],$

Wherein, y _das[n] is the output of traditional time delay superposition (delay-and-sum, DAS) Beam-former.

StS-CF can think to improve the one of CF.In like manner also can do similar process to symbol coherence factor (SCF), thus level and smooth symbol coherence factor (spatio-temporally smoothed sign coherence factor, StS-SCF) when obtaining sky.What CF utilized is the variance of signal amplitude, and SCF utilizes the standard deviation of the sign bit of Received signal strength to measure coherence.Therefore, the calculating of StS-SCF realizes through two steps:

First, by the amplitude x in StS-CF definition above _m[n+k] replaces with its sign bit b _m[n+k] (when amplitude is more than or equal to 0, its sign bit is+1, and when being less than 0, sign bit is-1), thus calculate an intermediate quantity StS-SCF ^*;

Then, StS-SCF is obtained by following formula:

$\mathrm{StS}-\mathrm{SCF}\left[n\right]=1-\sqrt{1-\mathrm{StS}-{\mathrm{SCF}}^{*}\left[n\right]}.$

Equally, by this factor pair signal and do weighting process.

In addition, based on proposed coherence factor class formation method, some approach can be used to the quality improving image further.Such as:

(1) StS-CF with StS-SCF is except doing except weighting process (being multiplied) to the output of DAS, can also do weighting process to the output of minimum variance (minimum variance, MV) Wave beam forming.Like this, the advantage of MV high spatial resolution can be obtained, the side lobe effects such as clutter can be suppressed again better and improve contrast further.

(2) each imaging point or the corresponding coherence factor value of time samples point.First can do space filtering (spatial filtering) smoothing processing to factor values, then filtered factor values is used for doing weighting process to the output of DAS or MV.The method can reduce the speckle variance of background area further, makes it more level and smooth, thus improves contrast-response characteristic.

(3) to put forward coherence factor method be a kind of method processed channel receiving signal, can combine with means such as aperture design.Such as, it can combine with space compound (spatialcompounding) imaging, speckle noise and clutter is suppressed simultaneously preferably, thus improves contrast further.

On Mathematical, coherence factor class methods can be received post filtering framework based on dimension and analyze.Weighter factor can be regarded as the dimension done the output of a given Beam-former and receive post filtering, that is:

$F_{\mathrm{wiener}}=\underset{F}{\arg \min }E\left\{{\left|A\right[n]-F{w}^{H}x[n\left]\right|}^2\right\}$

Wherein E represents and asks mathematic expectaion, and H represents conjugate transpose, and x [n] is the Received signal strength (vector form) of each passage after time delay, and w is the weight vectors of given Beam-former, and A [n] is desired signal.Received signal strength can regard desired signal and interference plus noise component sum as.

Time average process in StS-CF can be interpreted as a correction to above formula, namely considers that from the Received signal strength of an imaging point be the one section of pulse { x [t] } comprising multiple sample point, and is not only a sample point x [n].Like this, above formula can be revised as:

$F_{\mathrm{wiener}}=\underset{F}{\arg \min }E\left\{{\left|\right|\left\{A\left[t\right]\right\}-F\left\{w^{H}x\left[t\right]\right\}\left|\right|}_2^2\right\}$

$=\underset{F}{\arg \min }{\mathrm{\&Sigma;}}_{t=t_1}^{t_2}E\left\{{\left|A\right[t]-F{w}^{H}x[t\left]\right|}^2\right\}$

For convenience's sake, if the beginning and ending time sequence number of this section of pulse is t ₁=n-K, t ₂=n+K, pulse length is 2K+1.Here, suppose that A [t] is unknown and is deterministic, x [t] is random.A [t] needs suitable to measure generation with one, uses E{w ^hx [t] } provide.The analytic solutions of this minimization problem are:

$F_{\mathrm{wiener}}=\frac{{\mathrm{\&Sigma;}}_{t=t_1}^{t_2}{\left|A\right[t\left]\right|}^2}{{\mathrm{\&Sigma;}}_{t=t_1}^{t_2}E\left\{{\left|w^{H}x\right[t\left]\right|}^2\right\}}=\frac{{\mathrm{\&Sigma;}}_{t=t_1}^{t_2}{\left|E\right\{w^{H}x\left[t\right]\left\}\right|}^2}{{\mathrm{\&Sigma;}}_{t=t_1}^{t_2}E\left\{{\left|w^{H}x\right[t\left]\right|}^2\right\}},$

Wherein the Received signal strength sample by reality is estimated by mathematic expectaion.Search Space Smoothing can be used to estimate the molecule denominator of above formula, to obtain more accurate and healthy and strong estimator.As previously mentioned, it whole array (M array element) is divided into M-L+1 overlapping respectively containing the subarray of L array element, thus obtain one group and observe sample accordingly, and these samples are done on average estimate expectation.So when given Beam-former is the DAS of even weight (w is the vector of complete 1), the estimated result of the above formula obtained is exactly StS-CF.It is pointed out that based on different w(as DAS, MV), above formula can provide the definition of corresponding coherence factor.StS-CF based on DAS has less amount of calculation, is easier to realize in current ultrasonic image-forming system.

Particularly, the process of imaging and effect illustrate by emulation below and true experiment data instance.

(1) simulation example

It is the emulating image using different coherence factor method to obtain shown in Fig. 2.Wherein, Fig. 2 (a) is the cyst body mould emulating image obtained by traditional DAS Wave beam forming; Fig. 2 (b) is through the emulating image of this body mould after CF weighting process; Fig. 2 (c) is through the emulating image (p=1) of this body mould after SCF weighting process; Fig. 2 (d) is through the emulating image of this body mould after StS-CF (K=0, L=48) weighting process; Fig. 2 (e) is through the emulating image of this body mould after StS-CF (K=30, L=48) weighting process; Fig. 2 (f) is through the emulating image of this body mould after StS-SCF (K=0, L=48) weighting process; Fig. 2 (g) is through the emulating image of this body mould after StS-SCF (K=30, L=48) weighting process.

This experiment has been come by Field II ultra sonic imaging simulation software.Designed body mould comprises an echoless cyst and a high echo cyst.In 30 × 10 × 40 cubic millimeters of bodies, arrange 200 000 scattering points randomly, its scattering amplitude is Gauss distribution.The amplitude of echoless cyst inscattering point is set to 0, and outside scattering point amplitude average out to cyst in high echo cyst 40 times.The linear sensor array emulated has 96 array elements, and the spacing at adjacent array element center is half centre wavelength, and otch (kerf) is 0.05 millimeter.Array element height is 10 millimeters, and mid frequency and sampling frequency are set to 4 and 120MHz respectively.Experiment adopts fixed transmission to focus on (depth of focus is at 50 millimeters of places) and Dynamic receive focus, uses two cycle sinusoidal pulses as the sine pulse of pumping signal and the peaceful weighting of the two cycle Chinese as array element shock response.Experimental simulation tradition phased array imaging mode, sweep limits is 30 ° of sector regions containing 65 scanning lines.Before Wave beam forming, extra Gaussian noise (be 60dB relative to the SNR of Received signal strength) is added in channel receiving signal, to simulate actual environment.

For the reconstruction of each scanning line, sensor emission focusing pulse also receives echo-signal.After completing delay and focusing, calculate CF, SCF, StS-CF and StS-SCF coherence factor value at each time samples point place according to the amplitude of Received signal strength or sign bit information.They are used for relevant to channel receiving signal respectively and do weighting process, thus obtain respective Wave beam forming output.Subsequently, envelope detected is carried out, logarithmic compression to every bar scanning line, and coordinate scan conversion (containing bilinear interpolation) and image display.In Fig. 2, the Dynamic Announce scope of all images is 80dB.It should be noted that, CF and StS-CF can calculate based on the real signal received or its complex analytic signal.That adopt in this example is the latter.The multiple analytical form of channel receiving signal is by doing Hilbert transform to obtain to it.

As seen from Figure 2, CF and SCF can suppress the clutter clutter in echoless cyst effectively, but can produce artifact, and as image background regions speckle variance becomes large, the spot pattern of high echo cyst peripheral region suffers to destroy (black hole artifact).StS-CF and StS-SCF of the present invention can eliminate these artifacts preferably when maintenance suppresses clutter.And they become image to protect background texture pattern better relative to CF and SCF, reduce speckle variance thus improve the contrast of image.In addition, to StS-CF(StS-SCF) the situation of K=0 and K=30 compare, the time average effect of the latter can reduce spot intensity change, produce evenly background area.Here, 2K+1=61 time samples point is equivalent to an exomonental length.

Shown in Fig. 3 be based on the present invention carry StS-CF the emulating image that obtains of two kinds of corrective measures.Dynamic Announce scope is 80dB.The body modulus adopted is according to identical with Fig. 2.Fig. 3 (a) is after the coherence factor StS-CF value calculating each sample point place, does space filtering (spatial filtering) to it.In this example, the realization of filtering is simple 5 × 5 mean filters of use one (replacing the value of center by the meansigma methods of these 25 values).Filtered factor values is again to receiving the relevant of channel signal and (coherent sum) weighting.Can see, can reduce the speckle noise caused by the fluctuation of factor values to the Filtering Processing of coherence factor, thus speckle variance is reduced, contrast improves.But it also may slight degradation cyst (cyst) edge contour.

Fig. 3 (b) is image StS-CF Factors Weighting method and space compound (spatial compounding) imaging combined together obtained.Space compound realizes separate formed by same area-of-interest or that part is relevant multiple image by combining.These images are produced by the sensor in different spatial, or perhaps from different perspectives same area-of-interest observation imaging is obtained.In this example, by sensor (96 array element) translation successively of above-mentioned emulation, the spacing of each translation is the aperture length of 0.2 times, and translation obtains 6 width images 5 times altogether.In the forming process of every width image, StS-CF formation method is used (as mentioned above).Finally, 6 width images are done on average obtain final combination picture.Space compound can reduce speckle noise significantly, makes background area more evenly (reduction of speckle variance).From Fig. 3 (b), the present invention combined with space compound technology and can obtain the advantage of both simultaneously, that is, speckle noise and clutter clutter are suppressed (above-mentioned artifact phenomenon is also eliminated simultaneously) significantly.Contrast strengthen, picture quality can be further improved.

(2) truly data instance is tested

This example use University of Michigan before " geabr " data set of gathering of ultrasound in medicine and biology laboratory.It is the complete data set obtained by synthetic aperture focusing technology.Linear transducer has 64 array elements, array element distance 0.24 millimeter, mid frequency 3.333MHz, sampling frequency 17.76MHz.

It is the image using synthesis transmitting aperture technology (synthetic transmit aperturetechnique) to rebuild shown in Fig. 4.Wherein, Fig. 4 (a) is the image of the geabr experimental data obtained by traditional DAS Wave beam forming; Fig. 4 (b) is through the image of these data after CF weighting process; Fig. 4 (c) is through the image (p=1) of these data after SCF weighting process; Fig. 4 (d) is through the image of these data after StS-CF (K=22, L=32) weighting process; Fig. 4 (e) is through the image of these data after StS-CF (K=22, L=16) weighting process; Fig. 4 (f) is through the image of these data after StS-SCF (K=22, L=32) weighting process; Fig. 4 (g) is through the image of these data after StS-SCF (K=22, L=16) weighting process.

First passage RF signal carries out filtering and noise reduction through 64 rank finite impulse response band filters.This wave filter normalization cut-off frequency scope is [0.23 0.67], and is the triumphant damp window function windowing of 7 by beta value.For realizing more accurate focusing, before Applicative time time delay, 4 times of up-samplings are carried out to channel signal.Synthetic aperture is rebuild image process and is realized the two-way dynamic focusing of transmitting and receiving.Each transmitting (launching array element by one) all forms a width subimage.In subimage is rebuild, calculate the time delays of each imaging point relative to each reception array element, select accordingly also to calculate coherence factor value (CF, SCF, StS-CF and StS-SCF) corresponding to this imaging point closest to the sample of delay value from each Received signal strength, by factor values to the relevant of this some place Received signal strength sample and weighting to obtain the output of this point.Like this, the output of all imaging points forms a width subimage.All subimages are added the total output combining and form each imaging point.Carry out again below getting the processes such as envelope (simply, taking absolute value), scan conversion, logarithmic compression and image display.In Fig. 4, all image's dynamic display scopes are 60dB.

The image produced from CF and SCF shown in Fig. 4 (b) and (c), although the clutter in echoless cyst obtains good suppression, but image has serious artifact: the spot pattern that 1) overall brightness of image reduces background even is seriously removed, the especially signal to noise ratio lower ground side such as far-field region; 2) the brightness variance of background area speckle becomes large, makes background even not; 3) amplitude of the line target at far field place is difficult to observe by underestimation in the picture; 4) around high echo cyst, there is black hole artifact to produce.These phenomenons may affect diagnosis and detection clinically, make image be not suitable for medical imaging application.Background texture information may be lost.The contrast of echoless cyst reduces on the contrary.In the image that StS-CF and StS-SCF produces, above-mentioned artifact is removed significantly.The background area uniform smooth more of image.The contrast strengthen of echoless cyst, detectability improves.Wherein, StS-CF(K=22, L=M/2=32) image there is the most uniform spot pattern and minimum artifact.

In addition, this example also investigates the size of different subarray length L to the impact of imaging effect.Parameter L provides the balance between a Sidelobe Suppression performance and vigorousness.Less L(L=M/4) can clutter reduction clutter better, but also may cause more artifact.However, the method that the present invention carries gives a good counterbalance effect between both.The selection of L can be determined according to concrete clinical practice environment.Such as, when area-of-interest (ROI) is low echo target, L=M/2 is good selection, and when ROI is high echo target, L=M/4 may be better.

In a word, the present invention is a kind of corrective measure to traditional coherence factor class Adaptive beamformer ultrasonic imaging method.It can eliminate the artifact problem that coherence factor brings while maintaining secondary lobe and clutter suppression, its vigorousness is improved, thus is applicable to medical application better.Relative to traditional DAS Beamforming Method, it can strengthen the contrast of image, makes the spatial resolution of image also increase simultaneously.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.

Claims (7)

1. level and smooth coherence factor class adaptive ultrasound imaging method time empty, comprises the following steps:

(1) Wave beam forming pre-treatment;

(2) Wave beam forming process, comprise and delay and focusing is carried out to each channel receiving signal, the coherence factor value of level and smooth coherence factor class when calculating empty according to the channel receiving signal after delay process, utilize this coherence factor to received signal relevant and do weighting process, the Wave beam forming obtaining every bar scanning line exports;

(3) Wave beam forming post processing, comprises and carries out envelope detected, logarithmic compression, scan conversion and display to scanning line,

Wherein, level and smooth coherence factor (spatio-temporallysmoothed coherence factor, StS-CF) when described coherence factor is empty, is defined as follows:

$\mathrm{StS}-\mathrm{CF}\left[n\right]=\frac{{\mathrm{\&Sigma;}}_{k=-K}^K{\left|{\mathrm{\&Sigma;}}_{l=0}^{M-L}{\mathrm{\&Sigma;}}_{m=l}^{L+l-1}{x}_m\right[n+k\left]\right|}^2}{(M-L+1){\mathrm{\&Sigma;}}_{k=-K}^K{\mathrm{\&Sigma;}}_{l=0}^{M-L}{\left|{\mathrm{\&Sigma;}}_{m=l}^{L+l-1}{x}_m\right[n+k\left]\right|}^2}$

Wherein, M is the sum of receiving sensor array element; x _m[n+k] is the signal after time delay that m array element receives; N and n+k in square brackets is time index sequence number; L and K is two argument of type integers, and wherein L is user-defined parameter, and value is between 1 and M, and the ultra sonic imaging application according to reality is determined, the value of k is [-K, K], represents 2*K+1 time samples, or,

Level and smooth symbol coherence factor (spatio-temporallysmoothed sign coherence factor, StS-SCF) when described coherence factor is empty, is defined as follows:

$\mathrm{StS}-\mathrm{SC}{F}^{*}\left[n\right]=\frac{{\mathrm{\&Sigma;}}_{k=-K}^K{\left|{\mathrm{\&Sigma;}}_{l=0}^{M-L}{\mathrm{\&Sigma;}}_{m=l}^{L+l-1}{b}_m\right[n+k\left]\right|}^2}{(M-L+1){\mathrm{\&Sigma;}}_{k=-K}^K{\mathrm{\&Sigma;}}_{l=0}^{M-L}{\left|{\mathrm{\&Sigma;}}_{m=l}^{L+l-1}{b}_m\right[n+k\left]\right|}^2}$

$\mathrm{StS}-\mathrm{SCF}\left[n\right]=1-\sqrt{1-\mathrm{StS}-\mathrm{SC}{F}^{*}\left[n\right]}$

Wherein, M is the sum of receiving sensor array element; x _m[n+k] is the signal after time delay that m array element receives, b _m[n+k] is x _mthe sign bit of [n+k]; N and n+k in square brackets is time index sequence number; L and K is two argument of type integers, and wherein L is user-defined parameter, and value is between 1 and M, and the ultra sonic imaging application according to reality is determined, the value of k is [-K, K], represents 2*K+1 time samples.

2. method according to claim 1, wherein, described Wave beam forming pre-treatment comprises setting sonac transmitting and receiving pattern, carries out digitized, amplification, filtering to received signal.

3. method according to claim 1, wherein, when described coherence factor is empty during level and smooth coherence factor, L is taken as M/2.

4. method according to claim 1, wherein, when described coherence factor is empty during level and smooth coherence factor, utilize this coherence factor to received signal relevant and do weighting process, the output obtaining Wave beam forming is specially:

$y\left[n\right]=\mathrm{StS}-\mathrm{CF}\left[n\right]*y_{\mathrm{das}}\left[n\right]=\mathrm{StS}-\mathrm{CF}\left[n\right]*\frac{1}{M}{\mathrm{\&Sigma;}}_{m=0}^{M-1}{x}_m\left[n\right],$

Wherein, y _das[n] is the output of traditional time delay superposition (delay-and-sum, DAS) Beam-former.

5. method according to claim 1, wherein, when described coherence factor is empty during level and smooth symbol coherence factor, utilize this coherence factor to received signal relevant and do weighting process, the output obtaining Wave beam forming is specially:

$y\left[n\right]=\mathrm{StS}-\mathrm{SCF}\left[n\right]*y_{\mathrm{das}}\left[n\right]=\mathrm{StS}-\mathrm{SCF}\left[n\right]*\frac{1}{M}{\mathrm{\&Sigma;}}_{m=0}^{M-1}{b}_m\left[n\right],$

Wherein, y _das[n] is the output of traditional time delay superposition (delay-and-sum, DAS) Beam-former.

6. method according to claim 1, wherein, utilizes this coherence factor to received signal relevant and does weighting process and comprise and do weighting process to the output of minimum variance (minimum variance, MV) Wave beam forming.

7. method according to claim 1, wherein, described utilize this coherence factor to received signal relevant and do weighting process also comprise: first space filtering (spatial filtering) is carried out to described coherence factor value, then filtered coherence factor value is used for received signal relevant and do weighting process.