A method for collecting and processing micro blood flow in coronary arteries based on ultrasound

Abstract

The present invention provides a method for collecting and processing micro-blood flow of coronary arteries based on ultrasound, comprising the following steps: S1, obtaining a demodulated baseband channel data set; S2, obtaining a tissue signal and a blood flow signal; S3, executing a third processing path on the tissue signal, and simultaneously obtaining myocardial motion estimation information and a spectral Doppler image; S4, inputting the myocardial motion estimation information into a first processing path and a second processing path respectively as a feedback mechanism; S5, executing the first processing path on the tissue signal, and obtaining a grayscale tissue image in combination with the myocardial motion estimation information; S6, executing the second processing path on the blood flow signal, and obtaining a power Doppler image and a velocity Doppler image in combination with the myocardial motion estimation information; S7, outputting a grayscale tissue image, a power Doppler image, a velocity Doppler image and a spectral Doppler image as a final image. The present invention can realize high-sensitivity imaging of micro-vessels of coronary arteries without contrast agents.

Description

Technical Field

[0001] The present invention relates to the technical field of ultrasonic imaging, and in particular to a method for collecting and processing micro-blood flows in coronary arteries based on ultrasonic waves. Background Art

[0002] Coronary microvascular disease is one of the important causes of myocardial ischemia. Coronary microvascular disease refers to a clinical syndrome with objective evidence of exertional angina or myocardial ischemia caused by structural and / or functional abnormalities of precoronary arterioles and arterioles under the action of multiple pathogenic factors. Currently, common coronary angiography in medicine can only see large blood vessels with a diameter greater than 400-500μm, but more than 90% of the coronary microvessels in the entire coronary tree cannot be displayed by coronary angiography. Ultrasound imaging is widely used in clinical diagnosis because of its advantages of real-time fast, non-invasive and radiation-free. In particular, in recent years, the ultrafast Doppler blood flow imaging architecture with high frame rate acquisition and high parallel data processing has improved the quality of blood flow imaging and promoted the development of micro blood flow imaging.

[0003] Unfocused wave emission is a typical feature of ultrafast imaging, and its advantage is that it breaks through the imaging area and frame rate limitations of traditional ultrasound. Its basic framework is to construct a sound field that covers a larger area with a single emission, and use the received data obtained with fewer emissions to synthesize data frames with higher frame rates than traditional focusing methods. Later, a coherent compounding method was explored, which controls the deflection angle by delaying the emission to obtain low-resolution image frames at multiple angles, and then performs coherent compounding to obtain the final imaging result. However, in the context of ultrafast imaging, the number of unfocused waves must be kept as low as possible, and a small amount of wide beam emission will still sacrifice image resolution while obtaining a larger synthetic field and high frame rate, resulting in a lower signal-to-noise ratio and spatial resolution, thereby affecting the quality of Doppler microvascular imaging.

[0004] Although the method proposed in CN114533122A can effectively suppress noise and provide more robust microbubble positioning, it is not suitable for application in coronary microvascular imaging. Excessive microbubble size will affect the segmentation and visualization of microvessels. Although the method proposed in CN116269489A can effectively improve the imaging quality of microblood flow, it does not demodulate the channel data. The channel data has a high frequency band and a wide bandwidth, which consumes the working resources of the beam synthesis step; and it cannot simultaneously obtain tissue images, power and velocity blood flow images, and spectral Doppler images. The method proposed in CN117958864A can synchronously display multiple types of images, but it needs to separately transmit tissue excitation pulse signals and blood flow excitation pulse signals in time-sharing, and receive two groups of echo signals after excitation, which increases the amount of raw data collected and stored, and the control method is relatively complex, which also affects the synchronization. Summary of the invention

[0005] In view of this, the purpose of the present invention is to propose a method for collecting and processing cardiac coronary micro-blood flow based on ultrasound, so as to solve the problems that the current ultrasonic ultrafast micro-blood flow imaging based on non-focused waves has low signal-to-noise ratio and spatial resolution, the micro-blood flow extraction effect is not ideal, and it is impossible to obtain richer low-speed coronary artery microvascular images.

[0006] The technical means adopted by the present invention are as follows:

[0007] A method for collecting and processing micro blood flow in coronary arteries based on ultrasound comprises the following steps:

[0008] S1, collecting channel echo data of the human heart to be tested, performing orthogonal demodulation, spectrum shifting to baseband and downsampling processing on the channel echo data set formed by each channel echo data, and obtaining a demodulated baseband channel data set;

[0009] S2, performing clutter filtering on the baseband channel data set to obtain tissue signals and blood flow signals;

[0010] S3, executing a third processing path on the tissue signal, and obtaining myocardial motion estimation information and a spectral Doppler image at the same time;

[0011] S4, obtaining motion compensation information through myocardial motion estimation information, and inputting the motion compensation information into the first processing path and the second processing path respectively as a feedback mechanism;

[0012] S5, executing the first processing path on the tissue signal, and obtaining a grayscale tissue image in combination with the myocardial motion estimation information;

[0013] S6, executing a second processing path on the blood flow signal, and obtaining a power Doppler image and a velocity Doppler image in combination with the myocardial motion estimation information;

[0014] S7, outputting a grayscale tissue image, a power Doppler image, a velocity Doppler image and a spectrum Doppler image as final images.

[0015] Further, in S1, an ultrasonic signal is transmitted to the human heart to be tested through an ultrasonic transducer, and an ultrasonic echo signal carrying cardiac tissue information and hemodynamic information is obtained after being reflected by the human heart to be tested. The ultrasonic echo signal is then collected by the ultrasonic transducer to obtain channel echo data, which specifically includes the following steps:

[0016] Repeat multiple acquisitions of the target area of ​​the same human heart to be tested to obtain echo data sets of each channel at different deflection angles; directly perform orthogonal demodulation and downsampling processing on the echo data sets of each channel to obtain the demodulated baseband channel data set. The formula of the baseband channel data set is as follows:

[0017]

[0018] Where (x, z) is the horizontal and vertical pixel points in each emission, n is the nth channel of the ultrasonic transducer, and m is the mth deflection angle. is the channel echo data, is the angular frequency of the demodulated local oscillator signal.

[0019] Furthermore, in S2, the clutter filtering process uses a singular value decomposition filter to perform spatiotemporal clutter filtering on a baseband channel data set composed of a large set length, and divides the baseband channel data set into two parts: a tissue signal and a blood flow signal. The large set length refers to more than 100 frames.

[0020] Furthermore, S2 specifically includes the following steps:

[0021] The demodulated baseband channel dataset is of size A three-dimensional data matrix, where represents the axial depth, Indicates the number of beams received after beamforming, Indicates the length of the set;

[0022] Convert the three-dimensional data matrix into a two-dimensional space-time matrix through Casorati transformation S , the size of the two-dimensional space-time matrix S is , for the two-dimensional space-time matrix S Perform eigen-based singular value decomposition:

[0023]

[0024] Among them, △ is a diagonal matrix composed of singular values ​​sorted from large to small, U is the spatial singular vector matrix, V is the temporal singular vector matrix, represents the conjugate transpose of a matrix;

[0025] Calculate the covariance matrix ;

[0026] The singular value decomposition process is replaced by the eigenvalue decomposition process. The calculated time eigenvalues ​​and eigenvectors correspond to the singular values ​​△ and the time singular vector matrix V in the above formula. The eigenvalue decomposition calculation formula is as follows:

[0027]

[0028] in, represents the eigenvalue, E represents the corresponding eigenvector, and all eigenvalues ​​are arranged in descending order, that is, ;

[0029] Based on the eigenvalue decomposition calculation formula, the first-order derivative of the eigenvalue curve is calculated, and the eigenvalue corresponding to the maximum value of the first-order derivative is selected as the first threshold node k1 for separating tissue signals and blood flow signals, and the eigenvalue corresponding to the minimum value of the first-order derivative is selected as the second threshold node k2 for separating blood flow signals and noise. A set of eigenvalue groups representing tissue signals are generated respectively. and the corresponding blood flow signal to generate a set of characteristic value groups representing the blood flow signal ;

[0030] Use the feature vector group corresponding to the first group of tissue signal feature value groups The tissue signal can be extracted by filtering, and the eigenvector group corresponding to the eigenvalues ​​of the second group of blood flow signals is used. Filtering can extract the blood flow signal and obtain the tissue signal respectively. and blood flow signals :

[0031]

[0032] .

[0033] Furthermore, S5 specifically includes the following steps:

[0034] S51, beamforming the tissue signal: delaying the baseband channel data of several tissues after clutter filtering, adding the delayed data of each channel, and then performing phase rotation processing to obtain beamformed tissue signal data;

[0035] S52, the tissue signal data after beam synthesis is further processed by linear or nonlinear coherent compounding;

[0036] S53, performing envelope and logarithmic compression processing on the data set after coherent compounding;

[0037] S54, filtering the data processed in S53 to obtain a grayscale tissue image; the filtering process includes FIR filtering, smoothing filtering, and speckle noise suppression processing in the horizontal and vertical directions.

[0038] Furthermore, S6 specifically includes the following steps:

[0039] S61, the blood flow signal set is processed by beam synthesis: a plurality of blood flow related baseband channel data after clutter filtering are delayed, and the delayed data of each channel are added to obtain beam synthesized blood flow signal data;

[0040] S62, the blood flow signal data after beam synthesis is further processed by linear or nonlinear coherent compounding;

[0041] S63. Perform autocorrelation-based power estimation on the coherently compounded data set over the set length to obtain power Doppler data. The power Doppler data formula is as follows:

[0042]

[0043] Where r is the complex baseband signal after coherent recombination, is the complex conjugate of x, s is the spatial sample index, t is the temporal sample index, and T is the number of available samples, i.e., the set length;

[0044] Do velocity estimation based on autocorrelation to obtain velocity Doppler data. The velocity Doppler data formula is as follows:

[0045]

[0046] S64, filtering the power Doppler data to obtain a power Doppler image; filtering the velocity Doppler data to obtain a velocity Doppler image, wherein the filtering includes Kernel filtering and Tophat enhancement processing.

[0047] Furthermore, S3 specifically includes the following steps:

[0048] S31, beamforming the tissue signal: delaying the baseband channel data of several tissues after clutter filtering, adding the delayed data of each channel, and then performing phase rotation processing to obtain beamformed tissue signal data;

[0049] S32, re-sampling the tissue signal processed in S31, selecting S sample points representing the location of the myocardial tissue, obtaining all S sample points in each sampling, and accumulating all S sample points together to obtain another uniform data set based on the tissue signal;

[0050] S33, performing a fast Fourier transform on each sample point in the uniform data set over the set length, converting the time domain signal into a frequency domain signal, and obtaining the spectrum information of the uniform data set;

[0051] S34, performing motion estimation based on the frequency spectrum information of the uniform data set to obtain myocardial motion estimation information, wherein the myocardial motion estimation information includes a myocardial motion cycle, a peak velocity, and an average velocity;

[0052] S35. Perform short-time Fourier transform processing on the frequency spectrum information of the uniform data set to obtain a frequency spectrum Doppler image of the myocardial motion.

[0053] The present invention also provides a storage medium, which includes a stored program, wherein when the program is run, any of the above-mentioned methods for collecting and processing cardiac coronary micro-blood flows based on ultrasound is executed.

[0054] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes any one of the above-mentioned methods for collecting and processing ultrasonic coronary micro-blood flow through the computer program.

[0055] Compared with the prior art, the present invention has the following advantages:

[0056] The present invention can realize high-sensitivity imaging of coronary microvessels without contrast agent, effectively reduce the background interference of fast-moving myocardium through the feedback input mechanism of motion compensation information, and simultaneously obtain high signal-to-noise ratio, high-resolution grayscale cardiac tissue images, color blood flow images of myocardial coronary microvessels with richer blood flow information and better visualization, and spectral Doppler images reflecting myocardial motion. The anterior arterioles and arteriolar microvessels with a diameter of less than 400 μm will be effectively visualized.

[0057] Among them, directly performing orthogonal demodulation processing on the channel data set can move the channel signal from high frequency to baseband, obtain a baseband channel data set, and effectively reduce the channel data bandwidth; at the same time, beam synthesis based on baseband channel data has a lower phase error, which helps to improve the calculation accuracy, generate better beam synthesis data, and is beneficial to subsequent tissue and blood flow imaging. Further extraction of the baseband channel data can reduce redundant data and reduce the amount of calculation. Usually, the returned channel echo data is undemodulated radio frequency data. The radio frequency data is a real signal. The data frequency band is high, so a high sampling rate ADC is required to read, resulting in a wide transmission bandwidth of the channel data and a large amount of data. The present invention adopts orthogonal demodulation processing on the radio frequency channel data in advance, moves the spectrum to the baseband and performs downsampling processing, which can effectively reduce the channel data bandwidth; at the same time, beam synthesis based on the baseband channel data has a lower phase error, which helps to improve the calculation accuracy, generate better beam synthesis data, and is beneficial to subsequent tissue and blood flow imaging. Further extraction of the baseband channel data can reduce redundant data and reduce the amount of calculation.

[0058] In addition, clutter filtering is preferentially applied to the baseband channel data set before beamforming, which can effectively separate the data and obtain tissue signals for two-dimensional tissue imaging and blood flow signals carrying rich blood flow information for color blood flow imaging. At this time, the moving myocardial signals in the blood flow signals are suppressed in advance, reducing the interference of myocardial motion in subsequent color blood flow imaging, and avoiding the phase distortion of the blood flow signals in the subsequent nonlinear composite processing.

[0059] In processing baseband channel data sets, the present invention has low computational complexity and small amount of computation, effectively saves computational cost and improves computational speed, can quickly and effectively select appropriate boundary points of tissue signals and blood flow signals, blood flow signals and high-frequency noise, effectively separate data, and suppress part of high-frequency noise interference and myocardial tissue motion interference in advance, and respectively obtain tissue signals for two-dimensional tissue imaging and blood flow signals carrying rich blood flow information for color blood flow imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0061] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0063] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0064] like Figure 1 As shown, the present invention provides a method for collecting and processing micro blood flow in coronary arteries based on ultrasound, comprising the following steps:

[0065] S1, collecting channel echo data of the human heart to be tested, performing orthogonal demodulation and downsampling processing on the channel echo data set formed by each channel echo data, and obtaining a demodulated baseband channel data set;

[0066] An ultrasonic transducer transmits an ultrasonic signal (M non-focused waves with different deflection angles, including plane waves or diverging waves, usually M≥3) to the human heart to be tested. After reflection, the ultrasonic echo signal will carry the heart tissue information, and then the channel echo data will be obtained after being collected by the ultrasonic transducer. Repeat high frame rate acquisition of the same area of ​​interest to obtain a large set length, each channel echo data set at different deflection angles. In the data processing process, the collected echo data sets of each channel are first directly orthogonally demodulated and downsampled to obtain the demodulated baseband channel data set:

[0067]

[0068] Where (x, z) is the horizontal and vertical pixel points in each emission, n is the nth channel of the ultrasonic transducer, and m is the mth deflection angle. is the channel echo data, To demodulate the angular frequency of the local oscillator signal. Orthogonal demodulation processing can be implemented through hardware chips or software algorithms. Downsampling processing is to further extract the collected baseband channel data, reduce redundant data, and reduce the amount of data.

[0069] S2, performing clutter filtering on the baseband channel data set to obtain tissue signals and blood flow signals;

[0070] The baseband channel data set is subjected to clutter filtering. The clutter filtering mainly uses a singular value decomposition (SVD) filter to perform spatiotemporal clutter filtering on a three-dimensional data set composed of a large set length, and divides the baseband channel data set into two parts: tissue signal and blood flow signal.

[0071] The specific processing steps are as follows:

[0072] The demodulated baseband channel dataset is of size A three-dimensional data matrix, where represents the axial depth, Indicates the number of beams received after beamforming, In order to describe the spatiotemporal characteristics of the ultrasound data set, the three-dimensional ultrasound data matrix is ​​transformed into a two-dimensional spatiotemporal matrix S (whose size is ), perform feature-based singular value decomposition (SVD) on S:

[0073]

[0074] Among them, △ is a diagonal matrix composed of singular values ​​sorted from large to small, U is the spatial singular vector matrix, V is the temporal singular vector matrix, In order to further reduce the computational complexity, we first calculate the covariance matrix ; Then, the singular value decomposition process is replaced by the eigenvalue decomposition process, and the calculated time eigenvalues ​​and eigenvectors correspond to the singular values ​​△ and the time singular vector matrix V in the above formula. The eigenvalue decomposition calculation formula is as follows:

[0075]

[0076] in represents the eigenvalue, and E represents the corresponding eigenvector. All eigenvalues ​​are arranged in descending order, that is, According to the fact that tissue signals have the highest energy and spatial consistency, the eigenvectors corresponding to the larger eigenvalues ​​correspond to the tissue signals, the intermediate eigenvalues ​​are associated with the blood flow signals, and the smallest eigenvalues ​​are associated with the noise signals, so the blood flow signals can be described by the corresponding eigenvectors associated with the lower eigenvalues, and the tissue signals can be described by the corresponding eigenvectors associated with the higher eigenvalues.

[0077] Next, it is necessary to select appropriate eigenvalue nodes to distinguish tissue signals from blood flow signals. The specific selection principles of the present invention are as follows:

[0078] After obtaining the eigenvalue curve in descending order based on the above formula, the first-order derivative of the eigenvalue curve is calculated, and the eigenvalue corresponding to the maximum value of the first-order derivative (i.e., the steepest slope of the eigenvalue curve) is selected as the first threshold node k1 for separating tissue signals and blood flow signals, and the eigenvalue corresponding to the minimum value of the first-order derivative (i.e., the flattest point of the eigenvalue curve) is selected as the second threshold node k2 for separating blood flow signals and noise. At this time, a set of eigenvalue groups representing tissue signals are generated corresponding to tissue signals respectively. and the corresponding blood flow signal to generate a set of characteristic value groups representing the blood flow signal . Use the feature vector group corresponding to the first group of tissue signal feature value groups The tissue signal can be extracted by filtering, and the eigenvector group corresponding to the eigenvalues ​​of the second group of blood flow signals is used. The blood flow signal can be extracted by filtering. The tissue signal can be obtained by the following methods: and blood flow signals :

[0079]

[0080] S3, executing a third processing path on the tissue signal to obtain myocardial motion estimation information and a spectral Doppler image in sequence;

[0081] For the third processing path, the tissue signal obtained after the clutter filtering is resampled, and all S sample points under each sampling are obtained and accumulated together to obtain another set of uniform data sets based on tissue signals. Each point in the uniform data set is subjected to a fast Fourier transform on the set length, and the time domain signal is converted into a frequency domain signal to obtain the spectrum information of the data set. Motion estimation can be performed based on the spectrum set of all points to obtain myocardial motion estimation information, which is input into the first processing path and the second processing path as a feedback mechanism before filtering. It can be used as myocardial motion compensation information to effectively remove heartbeat background interference, reduce imaging artifacts, and help enhance the filtering effect, and improve the signal-to-noise ratio and resolution of myocardial coronary artery micro-blood flow imaging. In addition, the above spectrum set can also obtain a spectral Doppler image of myocardial motion.

[0082] S4, obtaining motion compensation information through myocardial motion estimation information, and inputting the motion compensation information into the first processing path and the second processing path as a feedback mechanism, specifically comprising the following steps:

[0083] Obtaining motion compensation information: After obtaining the spectrum information of the uniform data set, motion estimation is performed based on the spectrum information of the uniform data set, and myocardial motion estimation information is obtained by intuitively reading the spectrum set, including myocardial motion cycle, peak velocity and average velocity.

[0084] The motion compensation information is input into the first processing path and the second processing path: after the motion estimation information is obtained in the third processing path, in the filtering process of the first processing path, the filtering parameters are iteratively optimized based on the least squares method, and the motion estimation information is input as the optimization cutoff condition. When the iteration meets the cutoff condition, the final iteratively optimized filtering parameters are output, and then the next step of filtering process is carried out. The specific steps are as follows:

[0085] First, the initial value of the filter parameter is input, and the motion estimation information will be input as the optimization cutoff condition. The filter parameter is iteratively optimized based on the least squares method. The instant filter is generated based on the filter parameter value in each iteration process. The instant filter is used to filter and Fourier transform the data after envelope extraction and logarithmic compression processing in the first processing path to obtain a spectrum diagram. The period, peak speed and average speed information of the current filtered data are obtained according to the spectrum diagram. When the error between the period, peak speed and average speed information of the filtered data and the motion estimation information is minimized, the final iteratively optimized filter parameters are output. Then, the first processing path is filtered according to the optimized filter parameters to obtain myocardial motion data. The data after envelope extraction and logarithmic compression processing in the first processing path is subtracted from the myocardial motion, effectively removing the interference of the heartbeat background, reducing imaging artifacts, and obtaining the final grayscale tissue image. The feedback mechanism processing process of the second processing path is the same as that of the first processing path.

[0086] S5, executing the first processing path on the tissue signal, and obtaining a grayscale tissue image in combination with the myocardial motion estimation information;

[0087] For the first processing path, the tissue signal is first processed by beam synthesis. The beam synthesis processing here adopts the most common delay and sum (DAS) processing. The specific implementation method is: delay the baseband channel data related to several tissues after the clutter filtering, add the data after the delay processing of each channel, and then obtain the tissue signal data after beam synthesis after phase rotation processing. The tissue signal data after beam synthesis is further processed by coherent recombination (coherent recombination can be performed by direct linear superposition of each deflection angle, or nonlinear coherent recombination methods such as frame multiplication and summation (FMAS) can be used), and the envelope and logarithmic compression processing (logarithmic function with base 10, that is, ), and after filtering, a grayscale tissue image is obtained. The filtering process includes FIR filtering, smoothing filtering, and speckle noise suppression processing in the horizontal and vertical directions.

[0088] S6, executing a second processing path on the blood flow signal, and obtaining a power Doppler image and a velocity Doppler image in combination with the myocardial motion estimation information;

[0089] For the second processing path, the blood flow signal set obtained after clutter filtering is also processed by beam synthesis. The beam synthesis processing here also uses the most common delay and sum (DAS). The specific implementation method is: delay the baseband channel data related to several blood flows after clutter filtering, and add the data after delay processing of each channel to obtain the blood flow signal data after beam synthesis. After the blood flow signal data after beam synthesis is further processed by coherent compounding (coherent compounding can be performed by direct linear superposition of each deflection angle, or nonlinear coherent compounding methods such as frame multiplication and superposition (Frame-Multiply-And-Sum, FMAS)), the power Doppler data is obtained by performing autocorrelation-based power estimation on the set length for the coherent compounded data set:

[0090]

[0091] Where r is the complex baseband signal after coherent recombination, is the complex conjugate of x, s is the spatial sample index, t is the temporal sample index, and T is the number of available samples, i.e., the set length. At the same time, velocity Doppler data can be obtained by performing velocity estimation based on autocorrelation:

[0092]

[0093] After filtering, power Doppler images and velocity Doppler images are obtained respectively, wherein the filtering includes Kernel filtering and Tophat enhancement processing.

[0094] S7. When the final image is output, the grayscale tissue image of the coronary arteries of the heart, the color blood flow image of the microvessels (including the power Doppler image and the velocity Doppler image) and the spectrum Doppler image of the myocardial motion are synchronously displayed on the monitor.

[0095] The present invention belongs to the medical ultrasound industry and the field of ultrasound imaging technology. In this industry, it is specifically applied to the data collection, processing and imaging process of ultrasonic heart coronary artery micro-blood flow. With high-end ultrasound imaging equipment, it provides clinicians with clear and reliable high-quality images of human heart tissue grayscale, coronary artery microvessels, and Doppler spectrum. This method directly demodulates the collected stable, high signal-to-noise ratio, and high frame rate channel data, and uses a singular value decomposition clutter filtering algorithm to achieve fine signal decomposition, respectively obtain tissue signals, blood flow signals, and myocardial motion spectrum information, and process the three groups of data separately, and finally obtain ultrasonic heart coronary artery microvascular imaging with high signal-to-noise ratio and high spatial resolution, thereby improving the visualization effect.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for collecting and processing micro blood flow in coronary arteries based on ultrasound, characterized in that: The steps include: S1, collecting channel echo data of the human heart to be tested, performing orthogonal demodulation, spectrum shifting to baseband and downsampling processing on the channel echo data set formed by each channel echo data, and obtaining a demodulated baseband channel data set; S2, performing clutter filtering on the baseband channel data set to obtain tissue signals and blood flow signals; S3, executing a third processing path on the tissue signal, and obtaining myocardial motion estimation information and a spectral Doppler image at the same time; S31, beamforming the tissue signal: delaying the baseband channel data of several tissues after clutter filtering, adding the delayed data of each channel, and then performing phase rotation processing to obtain beamformed tissue signal data; S32, re-sampling the tissue signal processed in S31, selecting S sample points representing the location of the myocardial tissue, obtaining all S sample points in each sampling, and accumulating all S sample points together to obtain another uniform data set based on the tissue signal; S33, performing a fast Fourier transform on each sample point in the uniform data set over the set length, converting the time domain signal into a frequency domain signal, and obtaining the spectrum information of the uniform data set; S34, performing motion estimation based on the frequency spectrum information of the uniform data set to obtain myocardial motion estimation information, wherein the myocardial motion estimation information includes a myocardial motion cycle, a peak velocity, and an average velocity; S35, performing short-time Fourier transform processing on the frequency spectrum information of the uniform data set to obtain a frequency spectrum Doppler image of the myocardial motion; S4, obtaining motion compensation information through myocardial motion estimation information, and inputting the motion compensation information into the first processing path and the second processing path respectively as a feedback mechanism; S5, executing the first processing path on the tissue signal, and obtaining a grayscale tissue image in combination with the myocardial motion estimation information; S51, beamforming the tissue signal: delaying the baseband channel data of several tissues after clutter filtering, adding the delayed data of each channel, and then performing phase rotation processing to obtain beamformed tissue signal data; S52, the tissue signal data after beam synthesis is further processed by linear or nonlinear coherent compounding; S53, performing envelope and logarithmic compression processing on the data set after coherent compounding; S54, filtering the data processed in S53 to obtain a grayscale tissue image; the filtering process includes FIR filtering, smoothing filtering, and speckle noise suppression processing in the horizontal and vertical directions; S6, executing a second processing path on the blood flow signal, and obtaining a power Doppler image and a velocity Doppler image in combination with the myocardial motion estimation information; S61, the blood flow signal set is processed by beam synthesis: a plurality of blood flow related baseband channel data after clutter filtering are delayed, and the delayed data of each channel are added to obtain beam synthesized blood flow signal data; S62, the blood flow signal data after beam synthesis is further processed by linear or nonlinear coherent compounding; S63. Perform autocorrelation-based power estimation on the coherently compounded data set over the set length to obtain power Doppler data. The power Doppler data formula is as follows: Where r is the complex baseband signal after coherent recombination, r * yes r The complex conjugate of , s is the spatial sample index, t is the temporal sample index, and T is the number of available samples, i.e., the set length; Do velocity estimation based on autocorrelation to obtain velocity Doppler data. The velocity Doppler data formula is as follows: S64, filtering the power Doppler data to obtain a power Doppler image; filtering the velocity Doppler data to obtain a velocity Doppler image, wherein the filtering includes Kernel filtering and Tophat enhancement processing; S7, outputting a grayscale tissue image, a power Doppler image, a velocity Doppler image and a spectrum Doppler image as final images.

2. The method for collecting and processing micro blood flow in coronary arteries based on ultrasound according to claim 1, characterized in that: In S1, an ultrasonic signal is transmitted to the human heart to be tested through an ultrasonic transducer, and an ultrasonic echo signal carrying cardiac tissue information and hemodynamic information is obtained after being reflected by the human heart to be tested. The ultrasonic echo signal is then collected by the ultrasonic transducer to obtain channel echo data, which specifically includes the following steps: Repeat multiple acquisitions of the target area of ​​the same human heart to be tested to obtain echo data sets of each channel at different deflection angles; directly perform orthogonal demodulation and downsampling processing on the echo data sets of each channel to obtain the demodulated baseband channel data set. The formula of the baseband channel data set is as follows: Where (x, z) is the horizontal and vertical pixel points in each emission, n is the nth channel of the ultrasonic transducer, and m is the mth deflection angle. is the channel echo data, is the angular frequency of the demodulated local oscillator signal.

3. The method for collecting and processing micro blood flow in coronary arteries based on ultrasound according to claim 1, characterized in that: In S2, the clutter filtering process uses a singular value decomposition filter to perform spatiotemporal clutter filtering on a baseband channel data set composed of a large set length, and divides the baseband channel data set into two parts: tissue signal and blood flow signal. The large set length refers to more than 100 frames.

4. The method for collecting and processing micro blood flow in coronary arteries based on ultrasound according to claim 1, characterized in that: S2 specifically includes the following steps: The demodulated baseband channel dataset is of size A three-dimensional data matrix, where represents the axial depth, Indicates the number of beams received after beamforming, Indicates the length of the set; Convert the three-dimensional data matrix into a two-dimensional space-time matrix through Casorati transformation S , the size of the two-dimensional space-time matrix S is , for the two-dimensional space-time matrix S Perform eigen-based singular value decomposition: Among them, △ is a diagonal matrix composed of singular values ​​sorted from large to small, U is the spatial singular vector matrix, V is the temporal singular vector matrix, represents the conjugate transpose of a matrix; Calculate the covariance matrix ; The singular value decomposition process is replaced by the eigenvalue decomposition process. The calculated time eigenvalues ​​and eigenvectors correspond to the singular values ​​△ and the time singular vector matrix V in the above formula. The eigenvalue decomposition calculation formula is as follows: in, represents the eigenvalue, E represents the corresponding eigenvector, and all eigenvalues ​​are arranged in descending order, that is, ; Based on the eigenvalue decomposition calculation formula, the first-order derivative of the eigenvalue curve is calculated, and the eigenvalue corresponding to the maximum value of the first-order derivative is selected as the first threshold node k1 for separating tissue signals and blood flow signals, and the eigenvalue corresponding to the minimum value of the first-order derivative is selected as the second threshold node k2 for separating blood flow signals and noise. A set of eigenvalue groups representing tissue signals are generated respectively. and the corresponding blood flow signal to generate a set of characteristic value groups representing the blood flow signal ; Use the feature vector group corresponding to the first group of tissue signal feature value groups The tissue signal can be extracted by filtering, and the eigenvector group corresponding to the eigenvalues ​​of the second group of blood flow signals is used. Filtering can extract the blood flow signal and obtain the tissue signal respectively. and blood flow signals : 。 5. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is run, the ultrasound-based coronary micro-blood flow acquisition and processing method described in any one of claims 1 to 4 is executed.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the ultrasound-based coronary micro-blood flow acquisition and processing method as described in any one of claims 1 to 4 through the computer program.

Images