An improved ultrasound imaging method based on subarray coherence post-filtering

Abstract

The application discloses an improved ultrasonic imaging method based on subarray coherence post-filtering, which comprises the following steps: 1. transmitting ultrasonic waves to an imaging area and recording echo signals reflected back; 2. performing delay and superposition processing on the echo signals in a receiving aperture to obtain an imaging vector; 3. calculating a coherence factor, a normalized amplitude standard deviation reciprocal and a subarray length based on the imaging vector; 4. calculating a covariance matrix based on the imaging vector and the subarray length; 5. calculating a weight vector in a feature space-based minimum variance beamformer based on the covariance matrix; 6. calculating a subarray zero-crossing coherence factor based on the imaging vector; 7. calculating a post-filtering result based on adaptive subarray coherence; and 8. multiplying the post-filtering result with a superposition result of the imaging vector and imaging. The application can improve the contrast and resolution of an image.

Description

Technical Field

[0001] This invention belongs to the field of medical ultrasound imaging, specifically an improved ultrasound imaging method based on subarray coherence post-filtering. Background Technology

[0002] Medical ultrasound imaging equipment is one of the most common imaging devices, widely used in clinical practice due to its safety, convenience, and low cost. During medical ultrasound imaging, the ultrasound probe emits ultrasound waves into the human body and receives backscattered echoes from different tissue interfaces. By applying appropriate time delay compensation to the received signals of each array element, the echo signal vector of each imaging point can be obtained, and then superimposed using a delay-stacking (DAS) beamformer. Although this imaging method is simple, it suffers from low image quality.

[0003] To overcome the aforementioned shortcomings, adaptive beamformers have been extensively studied. Among them, the minimum variance (MV) beamformer significantly improves imaging resolution by adaptively calculating weighting coefficients by minimizing output power and applying distortion-free constraints. However, the MV beamformer has limited effectiveness in improving contrast. To address this, the eigenspace-based minimum variance (ESBMV) beamformer has been proposed. It divides the covariance matrix into signal and noise subspaces through eigenvalue decomposition and orthogonally projects the MV weights onto the signal subspace to suppress noise components, thereby effectively improving contrast. However, in low signal-to-noise ratio (SNR) environments, the echo signal is weak, and the eigenvalue distribution tends to be flat, making it difficult to threshold the signal and noise subspaces. If the threshold is set too high, some effective signals are mistakenly included in the noise subspace, resulting in dark spot artifacts; if the threshold is set too low, noise suppression is insufficient, limiting the imaging performance of ESBMV in low SNR scenarios.

[0004] On the other hand, coherence factor (CF) methods can effectively reduce sidelobes and clutter by weighting and optimizing the output of the distortionless beamformer. However, the impact of subarray decorrelation techniques on post-filter design has not been fully studied. To address this issue, some researchers have designed a subarray coherence based postfilter (SCBP) by utilizing the coherence of the subarray signals. However, in this post-filter, the noise power weight is fixed at the reciprocal of the subarray length, which cannot be dynamically adjusted according to the actual situation of the echo signal, thus limiting its improvement in resolution and contrast.

[0005] Therefore, it is essential to design an adaptive noise power weight based on the characteristics of the echo signal to overcome the performance limitations caused by the fixed weight and further improve the imaging quality. Summary of the Invention

[0006] The present invention addresses the shortcomings of the prior art by proposing an improved ultrasound imaging method based on subarray coherence post-filtering. By designing an adaptive noise power weight for subarray coherence post-filtering, noise and clutter signals can be effectively suppressed while the main lobe signal can be well preserved, thereby improving image contrast and resolution.

[0007] In order to achieve the above-mentioned objectives, the present invention proposes the following technical solutions: The improved ultrasound imaging method based on subarray coherence post-filtering of the present invention is characterized by the following steps: Step 1: During the process of the transducer emitting ultrasonic signals to the imaging area, acquire the backscattered echo signal reflected back from the imaging area; Step 2: Perform time-delay superposition processing on the backscattered echo signal at the receiving aperture to obtain the imaging vector. ,in, Indicates the first The echo signal corresponding to a plane wave at a given emission angle. It is the number of plane waves emitted; Step 3: Based on imaging vectors Calculate the coherence factor Standard deviation The reciprocal of the standard deviation and its normalized result Adaptive subarray length ; Step 4: Based on imaging vectors and length Calculate the covariance matrix after spatial smoothing and diagonal loading. ; Step 5: Perform eigenvalue decomposition on the covariance matrix and divide it into a signal subspace and a noise subspace. Project the minimum variance MV weights onto the signal subspace to obtain the weight vector. ; Step 6: Based on imaging vectors Calculate the zero-crossing coherence factor of the subarray ; Step 7: Based on imaging vectors Weight vector Normalized results Zero-crossing coherence factor of subarray Calculate the coherence of the subarray. Subarray incoherence and Adaptive noise power weighting Thus, the post-filtering result based on subarray coherence is obtained. ; Step 8: Transfer the imaging vector The superposition result and The results are multiplied, and then filtered, normalized, and logarithmically compressed to obtain high-quality ultrasound images.

[0008] The improved ultrasound imaging method based on subarray coherence post-filtering described in this invention is also characterized by step 3, which specifically includes the following steps: Step 3.1: Calculate the coherence factor using equation (1) : (1) Step 3.2: Calculate the standard deviation of the imaging vector. and the reciprocal of the standard deviation and to Normalization is performed to obtain the normalized result. ; Step 3.3: Calculate the length of the adaptive subarray using equation (2). : (2) In equation (2), This indicates rounding to the nearest integer. This represents the maximum subarray length, and its value ranges from 2 to... N / 2.

[0009] Furthermore, step 4 includes the following steps: Step 4.1: Calculate the covariance matrix using equation (3). : (3) In equation (3), Indicates the first Starting from a certain launch angle, continuously... The subvector formed by the echo signals corresponding to each emission angle. This represents the conjugate transpose operation; Step 4.2: Calculate the covariance matrix using equation (4). : (4) In equation (4), It is a load factor. Indicates size and The same identity matrix, and we have: (5) In equation (5), This indicates the computation of the trace of a matrix.

[0010] Furthermore, step 5 includes the following steps: Step 5.1: Use equation (6) to... Perform eigenvalue decomposition: (6) In equation (6), Represents the signal subspace. yes The corresponding eigenvalue matrix, Represents the noise subspace. yes The corresponding eigenvalue matrix; Indicates the threshold for partitioning, and , Represents the largest eigenvalue. For parameters, Indicates the first One eigenvalue; express The corresponding feature vector, This represents the conjugate transpose operation. ; Step 5.2: Project the minimum variance MV weights onto the signal subspace using equation (7) to obtain the weight vector. : (7) In equation (7), It is a length of The column vector, and Each element in the array is 1; yes The inverse matrix.

[0011] Furthermore, step 6 includes the following steps: Step 6.1: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require After multiplying the echo signals from adjacent angles, we obtain The product of the nth adjacent angles will be used to calculate the nth angle. The product of adjacent angles Compare with 0, if Then let the first Polarity flip flag bit Otherwise, let the first Polarity flip flag bit Thus, the polarity inversion flag bit vector is obtained. ; Step 6.2: Invert the polarity flag bit vector Starting from the beginning, move one element to the right each time to extract... A length of The continuous polarity flipping flag position vector; Step 6.3: Calculate the first step using equation (8). The sum of the polarity-flipped flag vectors Thus, the polarity reversal counting vector is obtained. : (8) In equation (8), Indicates the first A polarity-reversed flag position vector; Step 6.4: If Then let the first i Each binary flag bit Otherwise, let Thus, the binary flag bit vector is obtained. ; Step 6.5: Calculate the zero-crossing coherence factor of the subarray using equation (9). : (9).

[0012] Furthermore, step 7 includes the following steps: Step 7.1: Calculate the subarray coherence using equation (10). : (10) Step 7.2: Calculate the incoherence of the subarray using equation (11). : (11) Step 7.3: Calculate the adaptive noise power weights using equation (12). : (12) Step 7.4: Calculate the post-filtering result based on subarray coherence using equation (13). : (13) The present invention provides an electronic device, including a memory and a processor, characterized in that the memory is used to store a program supporting the processor in performing the method described therein, and the processor is configured to execute the program stored in the memory.

[0013] The present invention discloses a computer-readable storage medium storing a computer program, characterized in that the computer program is executed by a processor to perform the steps of the method described thereon.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention calculates the zero-crossing coherence factor of a subarray. An adaptive subarray length was designed to overcome the technical defects of fixed subarray length, which leads to a decrease in image contrast and resolution when the value is small, and a decrease in image speckle signal-to-noise ratio when the value is large. 2. Based on the echo signal characteristics of each imaging point, this invention adaptively determines the scaling factor for the coherent post-filtering of the subarray to dynamically balance noise suppression and signal fidelity, thereby improving the imaging quality of ultrasound images while enhancing the post-filter. 3. This invention adaptively calculates the scaling factor of the subarray coherent post-filter based on the dual characteristics of the amplitude statistics and phase coherence of the echo signal. Unlike single amplitude or single coherence characteristics, this invention uses a joint amplitude-phase decision, enabling better preservation of the main lobe signal and more effective suppression of off-axis clutter and noise, thereby improving image contrast and resolution. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention; Figure 2 These are the imaging results of different algorithms described in the specific embodiments of the present invention in simulation experiments; Figure 3 The images shown are transverse cross-sectional views of the imaging results of the point target at depths of 22 mm, 24 mm, and 26 mm using different algorithms in specific embodiments of the present invention. Detailed Implementation

[0016] In this embodiment, as Figure 1 As shown, an improved ultrasound imaging method based on subarray coherence post-filtering designs an adaptive noise power weight based on the amplitude and phase characteristics of the echo signal for use in the subarray coherence-based post-filtering. This post-filtering effectively suppresses noise and clutter signals while preserving the main lobe signal, thereby improving image contrast and resolution. Specifically, the method includes: Step 1: During the process of the transducer emitting ultrasonic signals to the imaging area, the backscattered echo signal reflected back from the imaging area is acquired; in this embodiment, Field II is used to simulate the ultrasonic scanning process in which a linear array transducer emits plane waves at different angles, and the plane waves are simultaneously received by all array elements after being reflected back by the scattering point.

[0017] Step 2: Perform time-delay superposition processing on the backscattered echo signal at the receiving aperture to obtain the imaging vector. ,in, Indicates the first The echo signal corresponding to a plane wave at a given emission angle. It represents the number of plane waves emitted.

[0018] Step 3: Based on imaging vectors Calculate the coherence factor Standard deviation The reciprocal of the standard deviation and its normalized result Adaptive subarray length .

[0019] Step 3.1: Calculate the coherence factor using equation (1) : (1) Step 3.2: Calculate the standard deviation of the imaging vector. and the reciprocal of the standard deviation and to Normalization is performed to obtain the normalized result. .

[0020] Step 3.3: Calculate the length of the adaptive subarray using equation (2). : (2) In equation (2), This indicates rounding to the nearest integer. This represents the maximum subarray length, and its value ranges from 2 to... N / 2.

[0021] In this embodiment, the subarray length used for covariance matrix estimation and zero-crossing coherence factor calculation is not a fixed value, but is adaptively determined according to the target type of the current imaging point. Specifically, a shorter subarray length is used for speckle regions to maintain a good speckle signal-to-noise ratio; a longer subarray length is used for main lobe, strong off-axis clutter, and incoherent noise regions to enhance clutter and noise suppression capabilities. Compared with a fixed subarray length, this invention avoids the following two types of problems caused by improper length selection: first, too short a length will weaken the ability to suppress noise and clutter, leading to a decrease in contrast and resolution; second, too long a length will reduce the speckle signal-to-noise ratio. Therefore, in this embodiment, by obtaining an adaptive subarray length based on the characteristics of the echo signal, a dynamic balance can be achieved between contrast, resolution, and speckle signal-to-noise ratio.

[0022] Step 4: Based on imaging vectors and length Calculate the covariance matrix after spatial smoothing and diagonal loading. ; Step 4.1: Calculate the covariance matrix using equation (3). : (3) In equation (3), Indicates the first i Starting from a certain launch angle, continuously... The subvector formed by the echo signals corresponding to each emission angle. This indicates the conjugate transpose operation.

[0023] Step 4.2: Calculate the covariance matrix using equation (4). : (4) In equation (4), It is a load factor. Indicates size and The same identity matrix, and we have: (5) In equation (5), This indicates the computation of the trace of a matrix.

[0024] Step 5: Perform eigenvalue decomposition on the covariance matrix and divide it into a signal subspace and a noise subspace. Project the minimum variance MV weights onto the signal subspace to obtain the weight vector. ; Step 5.1: Use equation (6) to... Perform eigenvalue decomposition: (6) In equation (6), Represents the signal subspace. yes The corresponding eigenvalue matrix, Represents the noise subspace. It is E n The corresponding eigenvalue matrix; Indicates the threshold for partitioning, and , Represents the largest eigenvalue. For parameters, Indicates the first One eigenvalue; express The corresponding feature vector, This represents the conjugate transpose operation; In this embodiment, .

[0025] Step 5.2: Project the minimum variance MV weights onto the signal subspace using equation (7) to obtain the weight vector. : (7) In equation (7), It is a length of The column vector, and Each element in the array is 1; yes The inverse matrix.

[0026] Step 6: Based on imaging vectors Calculate the zero-crossing coherence factor of the subarray ; Step 6.1: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require After multiplying the echo signals from adjacent angles, we obtain The product of adjacent angles , will the The product of adjacent angles Compare with 0, if Then let the first Polarity flip flag bit Otherwise, let the first Polarity flip flag bit Thus, the polarity inversion flag bit vector is obtained. .

[0027] In this embodiment, comparing the product of adjacent angles with 0 is to observe whether the polarity of the echo signal at adjacent angles has reversed. If the polarity has not reversed, the adjacent product value is greater than 0, and the polarity reversal flag is set to 0. Conversely, if the polarity has reversed, the adjacent product value is less than or equal to 0, and the polarity reversal flag is set to 1.

[0028] Step 6.2: Invert the polarity flag bit vector Starting from position 1, move one element to the right each time and extract... A length of The continuous polarity flipping flag position vector; Step 6.3: Calculate the first step using equation (8). i The sum of the polarity-flipped flag vectors Thus, the polarity reversal counting vector is obtained. : (8) In equation (8), Indicates the first A polarity-reversed flag position vector.

[0029] In this embodiment, the sum of the calculated polarity inversion flag sub-vectors ranges from 0 to... L- 1. If the sum of the calculated polarity reversal flag vectors is 0, it indicates that the number of polarity reversals is 0, and the polarity of the echo signals at adjacent angles is highly consistent; if the sum of the calculated polarity reversal flag vectors is larger, it indicates that the number of polarity reversals is more, and the phase distribution of the echo signals at adjacent angles tends to be more randomized.

[0030] Step 6.4: If Then let the first i Each binary flag bit Otherwise, let Thus, the binary flag bit vector is obtained. ; Step 6.5: Calculate the zero-crossing coherence factor of the subarray using equation (9). : (9) In this embodiment, the calculated It is located between 0 and 1. The higher the value, the more values ​​equal to 1 in the binary flag vector, the more values ​​whose sum of polarity inversion flag subvectors is 0, and the more coherent the signal; The lower the value, the fewer the values ​​of 1 in the binary flag vector, the fewer the values ​​of the polarity reversal flag subvectors whose sum is 0, and the less coherent the signal.

[0031] Step 7: Based on imaging vectors Weight vector Normalized results Zero-crossing coherence factor of subarray Calculate the coherence of the subarray. Subarray incoherence and Adaptive noise power weighting Thus, the post-filtering result based on subarray coherence is obtained. ; Step 7.1: Calculate the subarray coherence using equation (10). : (10) Step 7.2: Calculate the incoherence of the subarray using equation (11). : (11) Step 7.3: Calculate the adaptive noise power weights using equation (12). : (12) In this embodiment, The magnitude of the echo signal reflects the spatial dispersion of its amplitude. The larger the value, the more concentrated the amplitude of the echo signal and the smaller the fluctuations, meaning the energy is more concentrated in the direction of the main lobe. The smaller the value, the greater the dispersion of the echo signal amplitude and the more violent the fluctuations, meaning that the energy is dispersed in the side lobes or noise. The magnitude of reflects the phase consistency of the current imaging point in the angular domain. The larger the value, the more consistent the echo polarity and the more uniform the phase between different emission angles; The smaller the value, the more frequently polarity flips occur between adjacent angles, and the phase is randomly distributed. This invention uses... and The product of these two factors serves as a fundamental intermediate quantity, essentially constructing a dual verification logic. This logic applies if and only if the current imaging point simultaneously satisfies both amplitude concentration and phase regularity. Only when the value approaches 0 can it avoid misjudgment of a single feature in a specific scenario.

[0032] Step 7.4: Calculate the post-filtering result based on subarray coherence using equation (13). : (13) In this embodiment, It is a dynamic value calculated based on the amplitude and phase characteristics of the echo signal. When the value approaches 1, the signal is suppressed; when When the value approaches 0, the signal is output.

[0033] Step 8: Transfer the imaging vector The superposition result and The results are multiplied, and then filtered, normalized, and logarithmically compressed to obtain high-quality ultrasound images.

[0034] In this embodiment, an electronic device includes a memory and a processor. The memory stores a program that supports the processor in executing the above-described method, and the processor is configured to execute the program stored in the memory.

[0035] In this embodiment, a computer-readable storage medium stores a computer program, which is executed by a processor to perform the steps of the above method.

[0036] In this embodiment, the effects of the present invention are demonstrated through simulation experiments: The simulation experiment of a homogeneous medium was carried out in MATLAB using Field II simulation software. The specific implementation method is as follows: The simulation phantom used in this experiment included one anechoic cyst and three point targets. Specifically, the anechoic cyst had a radius of 2 mm and a center coordinate of (-5 mm, 24 mm). The coordinates of the three imaging point targets were (5 mm, 22 mm), (5 mm, 24 mm), and (5 mm, 26 mm), respectively. The longitudinal imaging range was between 20 and 28 mm, and the lateral imaging range was between -10 and 10 mm. Within the imaging region, scatterers were randomly distributed at a density of 10 scatterers per resolution unit, and each scatterer had a random amplitude. The imaging dynamic range was 60 dB. To verify the effectiveness of the algorithm, Gaussian white noise with an SNR of 20 dB was added to the echo data after beamforming at the receiving aperture in the simulation experiment. The array element center frequency used in the experiment was 5.208 MHz, the sampling frequency was 20 MHz, the number of array elements was 128, and the velocity of sound was 1540 m / s.

[0037] The obtained ultrasound image results are as follows Figure 2 As shown, the horizontal coordinate represents the lateral distance; the vertical coordinate represents the depth. In addition, in this embodiment, the ASSCBP method described in this invention and the Delayed Stacking Algorithm (DAS), based on subarray coherence post-filtering SCBP, were compared in an imaging experiment. The final imaging results are shown below. Figure 2 As shown, where Figure 2 Figure (a) in the figure represents the imaging results of DAS. Figure 2 Figure (b) in the image represents the imaging results of SCBP. Figure 2 Figure (c) in the diagram represents the imaging result of the ASSCBP method. Figure 2As shown, DAS produces the widest point target, and there is significant noise within the cyst. Compared to DAS, the noise within the cyst is somewhat suppressed in the SCBP method, and the point target is narrower, but noise still exists within the cyst, and the resolution improvement is limited. Compared to DAS and SCBP, the ASSCBP method described in this invention further suppresses noise within the cyst, and the point resolution is also significantly improved. In summary, the noise suppression capability and resolution improvement capability of this invention are optimal.

[0038] Figure 3 The graphs show a comparison of the lateral resolution of the three algorithms at depths of 22mm, 24mm, and 26mm for the scattering point target. The horizontal axis represents the lateral distance; the vertical axis represents the normalization magnitude; the black line represents the lateral resolution obtained using DAS, the blue line with a circle represents the lateral resolution obtained using SCBP, and the red line with a triangle represents the lateral resolution obtained using ASSCBP. Figure 3 Image (a) shows a comparison of the lateral resolution of the three algorithms when the scattering point target is at a depth of 22 mm. Figure 3 (b) in the figure is a comparison of the lateral resolution of the three algorithms when the scattering point target is at a depth of 24 mm; Figure 3 (c) in the figure is a comparison of the lateral resolution of the three algorithms when the scattering point target is at a depth of 26 mm. For example... Figure 3 As shown, the main lobe width obtained by the method ASSCBP at depths of 22mm, 24mm, and 26mm is the narrowest, indicating that the method ASSCBP can effectively improve the resolution compared to DAS and SCBP methods.

Claims

1. An improved ultrasound imaging method based on subarray coherence post-filtering, characterized in that, Includes the following steps: Step 1: During the process of the transducer emitting ultrasonic signals to the imaging area, acquire the backscattered echo signal reflected back from the imaging area; Step 2: Perform time-delay superposition processing on the backscattered echo signal at the receiving aperture to obtain the imaging vector. ,in, Indicates the first The echo signal corresponding to a plane wave at a given emission angle. It is the number of plane waves emitted; Step 3: Based on imaging vectors Calculate the coherence factor Standard deviation The reciprocal of the standard deviation and its normalized result Adaptive subarray length ; Step 4: Based on imaging vectors and length Calculate the covariance matrix after spatial smoothing and diagonal loading. ; Step 5: Perform eigenvalue decomposition on the covariance matrix and divide it into a signal subspace and a noise subspace. Project the minimum variance MV weights onto the signal subspace to obtain the weight vector. ; Step 6: Based on imaging vectors Calculate the zero-crossing coherence factor of the subarray ; Step 7: Based on imaging vectors Weight vector Normalized results Zero-crossing coherence factor of subarray Calculate the coherence of the subarray and Subarray incoherence and Adaptive noise power weighting Thus, the post-filtering result based on subarray coherence is obtained. ; Step 8: Transfer the imaging vector The superposition result and The results are multiplied and then filtered, normalized, and logarithmically compressed to obtain high-quality ultrasound images.

2. The improved ultrasound imaging method based on subarray coherence post-filtering according to claim 1, characterized in that, Step 3 specifically includes the following steps: Step 3.1: Calculate the coherence factor using equation (1) : (1) Step 3.2: Calculate the standard deviation of the imaging vector. and the reciprocal of the standard deviation and to Normalization is performed to obtain the normalized result. ; Step 3.3: Calculate the length of the adaptive subarray using equation (2). : (2) In equation (2), This indicates rounding to the nearest integer. This represents the maximum subarray length, and its value ranges from 2 to... N / 2.

3. The improved ultrasound imaging method based on subarray coherence post-filtering according to claim 1, characterized in that, Step 4 includes the following steps: Step 4.1: Calculate the covariance matrix using equation (3). : (3) In equation (3), Indicates the first Starting from a certain launch angle, continuously... The subvector formed by the echo signals corresponding to each emission angle. This represents the conjugate transpose operation; Step 4.2: Calculate the covariance matrix using equation (4). : (4) In equation (4), It is a load factor. Indicates size and The same identity matrix, and we have: (5) In equation (5), This indicates the computation of the trace of a matrix.

4. An improved ultrasound imaging method based on subarray coherence post-filtering according to claim 1, characterized in that, Step 5 includes the following steps: Step 5.1: Use equation (6) to... Perform eigenvalue decomposition: (6) In equation (6), Represents the signal subspace. yes The corresponding eigenvalue matrix, Represents the noise subspace. yes The corresponding eigenvalue matrix; Indicates the threshold for partitioning, and , Represents the largest eigenvalue. For parameters, Indicates the first One eigenvalue; express The corresponding feature vector, This represents the conjugate transpose operation. ; Step 5.2: Project the minimum variance MV weights onto the signal subspace using equation (7) to obtain the weight vector. : (7) In equation (7), It is a length of The column vector, and Each element in the array is 1; yes The inverse matrix.

5. An improved ultrasound imaging method based on subarray coherence post-filtering according to claim 1, characterized in that, Step 6 includes the following steps: Step 6.1: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require After multiplying the echo signals from adjacent angles, we obtain The product of the nth adjacent angles will be used to calculate the nth angle. The product of adjacent angles Compare with 0, if Then let the first Polarity flip flag bit Otherwise, let the first Polarity flip flag bit Thus, the polarity inversion flag bit vector is obtained. ; Step 6.2: Invert the polarity flag bit vector Starting from the beginning, move one element to the right each time to extract... A length of The continuous polarity flipped flag position vector; Step 6.3: Calculate the first step using equation (8). The sum of the polarity-flipped flag vectors Thus, the polarity reversal counting vector is obtained. : (8) In equation (8), Indicates the first A polarity-reversed flag position vector; Step 6.4: If Then let the first i Each binary flag bit Otherwise, let Thus, the binary flag bit vector is obtained. ; Step 6.5: Calculate the zero-crossing coherence factor of the subarray using equation (9). : (9)。 6. An improved ultrasound imaging method based on subarray coherence post-filtering according to claim 1, characterized in that, Step 7 includes the following steps: Step 7.1: Calculate the subarray coherence using equation (10). : (10) Step 7.2: Calculate the incoherence of the subarray using equation (11). : (11) Step 7.3: Calculate the adaptive noise power weights using equation (12). : (12) Step 7.4: Calculate the post-filtering result based on subarray coherence using equation (13). : (13)。 7. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store a program that supports a processor in executing the method of any one of claims 1-6, the processor being configured to execute the program stored in the memory.

8. A computer-readable storage medium storing a computer program thereon, characterized in that, The computer program is executed by a processor to perform the steps of the method according to any one of claims 1-6.

Images