Method and device for estimating motion of a chamber wall based on an adaptive regularized optical flow model

Abstract

The application relates to a method and device for estimating a ventricular wall motion based on an adaptive regularization optical flow model, belonging to the technical field of ultrasonic image processing. The method is characterized in that: firstly, the obtained optical flow vector is analyzed through two-dimensional wavelet analysis and function mapping to obtain motion boundary information of the motion field relative to sudden change; secondly, the sudden change motion boundary information is adaptively fed back to the optical flow calculation model to correct the regularization coefficient of the smoothing term and recalculate the optical flow field; finally, the optical flow calculation process is repeated until the wavelet high-frequency component of the optical flow field converges; the method reflects the gradient difference between the myocardial ventricular wall and the surrounding tissue in the regularization coefficient, thereby improving the accuracy of motion estimation.

Description

Technical Field

[0001] This application relates to the field of ultrasonic image processing technology, and in particular to a method and apparatus for estimating the motion of the chamber wall based on an adaptive regularized optical flow model. Background Technology

[0002] Cardiovascular diseases are a major threat to human health, ranking first among causes of death worldwide, with ischemic heart disease accounting for the largest proportion. Ischemic heart disease is a clinical condition mainly caused by insufficient blood and oxygen supply to the myocardium, leading to myocardial ischemia. Abnormal myocardial function can cause heart failure, myocardial infarction, and even sudden death. Therefore, effective assessment and real-time analysis of myocardial function are particularly important.

[0003] Currently, ultrasound technology, with its unique advantages of being non-invasive, real-time, and inexpensive, is widely used in clinical practice. Echocardiography can effectively display a patient's physiological and cardiac function, cardiac anatomy, and measure the patient's left ventricular diastolic and systolic capacity, effectively diagnosing conditions such as mitral regurgitation and left ventricular aneurysm. In recent years, intracardiac echocardiography has been applied as a new imaging method in routine examinations and intraoperative monitoring, such as left atrial appendage closure and catheter stent implantation. Echocardiography plays a crucial role in the diagnosis and analysis of cardiovascular diseases; however, diagnosis largely depends on the physician's subjective judgment, and differences in clinical experience can easily lead to incorrect diagnoses. Quantitative analysis of echocardiographic indicators can provide more accurate technical means for assessing myocardial function. In particular, through ventricular wall motion analysis, reliable quantitative indicators can be provided to clinicians, potentially reducing the risk of human error. However, due to the physical limitations of ultrasound imaging, factors such as speckle and noise often appear in the images, resulting in shadows and unclear myocardial wall contours in echocardiograms, which poses certain difficulties for calculating ventricular wall motion and affects its application value.

[0004] Utilizing the boundary information of the ventricular wall and its inner and outer surfaces provided by manual segmentation can improve the quality of motion estimation in echocardiography and aid in calculating displacement between adjacent frames. However, echocardiograms contain many irregular granular spots, and even within the ventricular wall, the texture information remains rich, resulting in significant noise and artifacts. Furthermore, changes in the doctor's probe placement and patient position can cause "shadowing" of the ventricular wall. These factors contribute to discrepancies between the results of manual segmentation and the true boundaries, and manual segmentation is inefficient for large volumes of ultrasound data. Summary of the Invention

[0005] This application provides a method and apparatus for estimating ventricular wall motion based on an adaptive regularized optical flow model. It automatically obtains the motion boundary information between the myocardial ventricular wall and surrounding tissues using wavelet analysis, and establishes an adaptive regularization parameter iteration process through feedback to gradually optimize the optical flow calculation model until the wavelet coefficients converge, thereby improving the accuracy of ventricular wall motion estimation in echocardiography.

[0006] The technical solution of this application is as follows:

[0007] According to a first aspect of the embodiments of this application, a method for estimating wall motion based on an adaptive regularized optical flow model is provided, comprising:

[0008] Step 101: Obtain a sorted combination of multiple ultrasound images within a preset cardiac cycle based on the ultrasound echocardiogram.

[0009] Step 102: Calculate the motion optical flow field of the two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method;

[0010] Step 103: Use two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of the two consecutive frames and extract motion boundary information;

[0011] Step 104: The motion boundary information is mapped and added to the regularization coefficient of the optical flow model to calculate the motion boundary smoothing coefficient;

[0012] Step 105: Recalculate the motion optical flow field of the two frames based on the motion boundary smoothing coefficient;

[0013] Step 106: Repeat steps 103, 104 and 105 until the wavelet coefficients converge to obtain the final estimate of the wall motion displacement.

[0014] Optionally, the step of calculating the motion optical flow field of the two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method includes:

[0015] The motion optical flow field of the two consecutive frames of ultrasound images is obtained by minimizing the total energy function of the two consecutive frames in the sorted combination of the multi-frame ultrasound images. The total energy function of the two consecutive frames is: E gradient and E color Adding them together represents the data items, E smooth This represents the smoothing term, where α is a regularization smoothing coefficient that is greater than zero.

[0016] Optional, data item E color According to the formula Calculate, data item E gradient According to the formula Calculate the smoothing term E.smooth According to the formula Calculate, where, ε is a constant, x represents the position of pixel A, and w represents the displacement of pixel A. This indicates the calculation of the derivative in the XY direction.

[0017] Optionally, the regularization smoothing coefficient α is determined using the formula... Calculate, where α g For global information weights, α l For local information weights, I x I y This represents the image gradient.

[0018] Optionally, the step of using two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of two consecutive frames of images and extract motion boundary information includes:

[0019] According to formula f j =f j-1 +cmpt uses two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of two consecutive frames and extract motion boundary information, where f j The highest approximation coefficients of the original signal f are derived from f. j It begins to decompose into lower-level approximations f. j-1 And wavelet component cmpt.

[0020] Optionally, the high-frequency component cmpt4 in the wavelet component cmpt is taken and obtained as the non-boundary smoothing coefficient W in the optical flow estimation through function mapping. cmpt According to the formula +W cmpt The motion boundary information is mapped and added to the regularization coefficient of the optical flow model to calculate the motion boundary smoothness coefficient.

[0021] Optionally, the steps 103, 104, and 105 are executed iteratively until the wavelet coefficients converge to obtain the final estimation result of the ventricular wall motion displacement, including:

[0022] Steps 103, 104, and 105 are executed iteratively until the difference between two adjacent wavelet high-frequency components converges, yielding the final wall motion displacement estimation result. According to a second aspect of the embodiments of this application, a wall motion estimation device based on an adaptive regularized optical flow model is provided, comprising:

[0023] The image processing module is used to obtain a sorted combination of multiple ultrasound images within a preset cardiac cycle based on the ultrasound cardiac images;

[0024] The first calculation module is used to calculate the motion optical flow field of the two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method.

[0025] The decomposition and reconstruction module is used to decompose and reconstruct the motion optical flow field of two consecutive frames of images using two-dimensional wavelet analysis and extract motion boundary information.

[0026] The mapping module is used to map the motion boundary information into the regularization coefficient of the optical flow model and calculate the motion boundary smoothing coefficient.

[0027] The second calculation module is used to recalculate the motion optical flow field of the two frames of images based on the motion boundary smoothing coefficient.

[0028] The processing module iteratively executes the decomposition and reconstruction module, the mapping module, and the second calculation module until the wavelet coefficients converge, thus obtaining the final estimation result of the wall motion displacement.

[0029] According to a third aspect of the embodiments of this application, a non-volatile storage device is provided, comprising: a processor, and a memory communicatively connected to the processor;

[0030] The memory stores computer-executed instructions;

[0031] The processor executes computer execution instructions stored in the memory to implement the method provided in the first aspect.

[0032] According to a fourth aspect of the embodiments of this application, a computer-readable storage medium is provided, wherein computer-executable instructions are stored in the computer-readable storage medium, and the computer-executable instructions are executed by a processor to implement the method provided in the first aspect.

[0033] According to a fifth aspect of the embodiments of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided in the first aspect.

[0034] Beneficial effects:

[0035] This application involves

[0036] A method and apparatus for estimating ventricular wall motion based on an adaptive regularized optical flow model are proposed. Addressing the problem that commonly used optical flow model estimation methods employ constant regularization smoothing coefficients, leading to large calculation errors in the ventricular wall boundary region and unclear motion boundaries, this invention proposes using wavelet analysis to construct an adaptive optical flow regularized model for ventricular wall motion estimation. First, the initially obtained optical flow vector is analyzed using two-dimensional wavelet analysis and function mapping to obtain motion boundary information with relatively abrupt changes in the motion field. Second, this abrupt motion boundary information is adaptively fed back into the optical flow calculation model to correct the regularization coefficient of the smoothing term and recalculate the optical flow field. Finally, the above optical flow calculation process is repeated until the high-frequency wavelet components of the optical flow field converge. This method reflects the gradient difference between the myocardial ventricular wall and surrounding tissues in the regularization coefficient, thereby improving the accuracy of motion estimation.

[0037] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0038] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application, and do not constitute an undue limitation of this application.

[0039] Figure 1 This is a flowchart illustrating an exemplary embodiment of a method for estimating wall motion based on an adaptive regularized optical flow model provided in this application.

[0040] Figure 2 This is an algorithm flowchart illustrating a method for estimating wall motion based on an adaptive regularized optical flow model, according to an exemplary embodiment.

[0041] Figure 3 This is a schematic diagram illustrating motion boundary extraction based on wavelet decomposition according to an exemplary embodiment;

[0042] Figure 4 This is a schematic diagram illustrating four components from low frequency to high frequency obtained by wavelet decomposition of an optical flow vector according to an exemplary embodiment.

[0043] Figure 5 This is a schematic diagram comparing the wall motion estimation results calculated according to an exemplary embodiment with the prior art;

[0044] Figure 6 This is a schematic diagram of a wall motion estimation device based on an adaptive regularized optical flow model, according to an exemplary embodiment. Detailed Implementation

[0045] To enable those skilled in the art to better understand the technical solutions of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0046] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0047] To improve image quality and obtain more accurate wall motion mechanics parameters for more effective assessment of myocardial function, this application collected echocardiographic images from 15 patients with abnormal myocardial motion and 15 normal patients. Echocardiographic wall motion estimation is of great significance for the diagnosis of heart disease. However, commonly used optical flow model estimation methods, which employ constant regularization smoothing coefficients, lead to large calculation errors in the wall boundary region and unclear motion boundaries. This invention proposes an adaptive optical flow regularization model for wall motion estimation using wavelet analysis. First, the initially obtained optical flow vector is analyzed using two-dimensional wavelet analysis and function mapping to obtain relatively abrupt motion boundary information of the motion field. Second, this abrupt motion boundary information is adaptively fed back into the optical flow calculation model to correct the smoothing term regularization coefficient and recalculate the optical flow field. Finally, the above optical flow calculation process is repeated until the high-frequency wavelet components of the optical flow field converge. This method reflects the gradient difference between the myocardial ventricular wall and surrounding tissues in the regularization coefficient, thereby improving the accuracy of motion estimation.

[0048] Figure 1 This document provides a flowchart of a ventricular wall motion estimation method based on an adaptive regularized optical flow model, as an exemplary embodiment of this application. The ventricular wall motion estimation method of this embodiment automatically obtains the motion boundary information between the myocardial ventricular wall and surrounding tissues using wavelet analysis. An adaptive feedback-based iterative process for regularized parameters is established to progressively optimize the optical flow calculation model until the wavelet coefficients converge, thereby improving the motion estimation accuracy of echocardiography. Figure 1 As shown, the specific steps of this method for estimating wall motion based on an adaptive regularized optical flow model are as follows:

[0049] Step 101: Obtain a sorted combination of ultrasound images within a cardiac cycle based on the ultrasound echocardiogram.

[0050] This invention uses echocardiographic images from 30 patients as raw image data, extracts and analyzes ultrasound images from one cardiac cycle, and calculates the displacement data of the myocardial ventricular wall motion. This has significant reference value for the diagnosis of myocardial ischemia in cardiovascular diseases. (Reference) Figure 2 As shown, the original image data can be preprocessed first, and then ultrasound images within a cardiac cycle can be extracted, sorted, and combined for analysis.

[0051] Step 102: Calculate the motion optical flow field of the two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method.

[0052] Whether myocardial function is abnormal can be evaluated by analyzing the coordination of myocardial ventricular wall motion in echocardiography. Motion estimation provides clinicians with quantitative indicators to aid in clinical diagnosis. The advantages of optical flow methods, such as multi-scale and multi-resolution capabilities, make accurate local motion analysis in echocardiography possible. The Brox optical flow method combines three assumptions and studies the energy equation for calculating optical flow: the constant grayscale assumption, the constant gradient assumption, and the constraint of maintaining spatiotemporal smoothness despite discontinuities. The constraints are data constraints and smoothing constraints, where the data constraints include the constant grayscale constraint and the constant gradient constraint. The total energy function is a weighted sum of the data and smoothing terms. Minimizing the total energy function and establishing a numerical solution method, i.e., finding target values ​​u and v that minimize the total energy function, yields the motion optical flow field for consecutive frames.

[0053] This invention employs a numerical solution to minimize the total energy function of two consecutive ultrasound images obtained from a sequence of multiple ultrasound images, thus calculating the motion optical flow field of the two consecutive images. The total energy function of the two consecutive images is: E gradient and E color Adding them together represents the data items, E smooth This represents the smoothing term, where α is a regularization smoothing coefficient greater than zero. Data term E color According to the formula Calculate, data item E gradient According to the formula Calculate the smoothing term E. smooth According to the formula Calculate, where, ε is a constant, x represents the position of pixel A, and w represents the displacement of pixel A. This indicates the calculation of the derivative in the XY direction.

[0054] The regularization smoothing coefficient α in this embodiment of the invention adopts the formula Calculate, where α g For global information weights, α l For local information weights, I xI y The gradient is the image gradient. The regularization coefficient α is inversely proportional to the gradient at that pixel, indicating that the regularization coefficient α is smaller at the boundaries of the ultrasound image or in areas with rich texture, and larger in areas with uniform grayscale.

[0055] Step 103: Use two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of the two frames and extract motion boundary information.

[0056] Utilizing the boundary information of the myocardial ventricular wall provided by segmentation can improve the quality of motion estimation in echocardiography and increase the accuracy of displacement calculation between adjacent frames. This invention utilizes wavelet transform to automatically extract motion information of the ventricular wall boundaries. By decomposing the signal into sub-signals of different frequency bands—that is, decomposing the signal into components located at different times and frequencies—the low-frequency sub-image becomes the brightness image, while the horizontal, vertical, and diagonal high-frequency sub-images become the detail images. To address the issue of blurred motion boundaries in the optical flow field during optical flow regularization, this invention uses the optical flow motion vector as the original signal for wavelet analysis, which can effectively extract the boundary information between the myocardial ventricular wall and surrounding tissues.

[0057] refer to Figure 2 and Figure 3 As shown, the embodiments of the present invention are based on formula f j =f j-1 +cmpt uses two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of two consecutive frames and extract motion boundary information, where f j These are the highest approximation coefficients of the original signal f. From f j It begins to decompose into lower-level approximations f. j-1 The wavelet component cmpt1 is used, and this decomposition process is repeated until the four wavelet components cmpt1, cmpt2, cmpt3, and cmpt4 are extracted. Then, the extracted high-frequency component cmpt4 is used to obtain the non-boundary smoothing coefficient Wcmpt in the optical flow estimation through function mapping.

[0058] Step 104: The motion boundary information is mapped and added to the regularization coefficient of the optical flow model to calculate the motion boundary smoothing coefficient.

[0059] In this embodiment of the invention, the high-frequency component cmpt4 in the wavelet component cmpt is obtained by function mapping to obtain the non-boundary smoothing coefficient W in optical flow estimation. cmpt According to the formula +W cmpt The motion boundary information is mapped and added to the regularization coefficient of the optical flow model to calculate the motion boundary smoothness coefficient.

[0060] Step 105: Recalculate the motion optical flow field of the two frames based on the motion boundary smoothing coefficient.

[0061] Step 106: Repeat steps 103, 104 and 105 until the wavelet coefficients converge to obtain the final estimate of the wall motion displacement.

[0062] Due to speckle and noise phenomena in echocardiography, the non-boundary regions of the myocardium have rich texture information and large gradient differences. This means that the regularization coefficient of the original optical flow method cannot effectively distinguish the myocardial ventricular wall from surrounding tissues. In this embodiment of the invention, a non-boundary smoothing coefficient W obtained from wavelet analysis is added to the regularization coefficient α. cmpt The extracted motion boundary information is used to adjust the regularization coefficients of the boundaries between the myocardial ventricular wall and surrounding tissues, as well as the coefficients within these boundaries; that is, reducing the coefficients in the boundary region and increasing the coefficients in the non-boundary region. In this embodiment, high-frequency coefficients obtained from wavelet decomposition are used to describe the boundary distribution, and the boundary distribution and non-boundary smoothing terms are correlated and added to the regularization coefficients. The regularization coefficient α in the optical flow calculation is calculated using the formula... +W cmpt Perform iterative calculations.

[0063] Among them W cmpt The smoothing coefficients are obtained through wavelet analysis and function mapping, specifically the smoothing coefficients W obtained by function mapping using the extracted high-frequency component cmpt4. cmpt The wavelet components are incorporated into the optical flow regularization process, iteratively updating the calculated optical flow results until the wavelet components converge. This process continues until the non-boundary smoothing coefficient W, obtained through wavelet analysis and function mapping, is reached. cmpt =a w* The final estimation result of the wall motion displacement is obtained, where aw is a constant. Since the extreme values ​​of wavelet coefficients can differ in sign due to the relative stiffness relationships between different tissues, in W... cmpt =a w* The absolute value is calculated in the correspondence, and the magnitude of the absolute value reflects the degree of smoothness required.

[0064] The embodiments of the present invention use formulas +W cmpt After iterative calculation to obtain the new regularization coefficient α2 to replace α1, a new optical flow motion field is calculated. The calculation result is then subjected to wavelet decomposition to obtain a new high-frequency component cmpt4'. This component is then used for secondary mapping to obtain a new smoothing coefficient W. cmpt2 This is then added to the regularization process to obtain a new regularization coefficient α3. After that, the optical flow calculation is performed again, and the above steps are repeated until the high-frequency component cmpt4 obtained by wavelet decomposition converges, thus obtaining the final displacement estimation result.

[0065] refer to Figure 4 As shown, to simulate myocardial motion, the wall motion estimation method of this invention uses a U-shaped binary image to simulate myocardial contraction. An initial optical flow vector is obtained by comparing two frames using the fixed smoothing coefficient optical flow method. This initial optical flow vector is then subjected to two-dimensional wavelet analysis to obtain wavelet components at four frequencies. The experimental results are shown below. Figure 4 As shown in the results, the high-frequency component (d) best reflects the motion boundary information of the optical flow vector and has the least noise. Figure 4 In the image, (a) is a U-shaped binary image, and (b), (c), (d), and (e) are the four components from low frequency to high frequency obtained after wavelet decomposition of the optical flow vector.

[0066] refer to Figure 5 As shown, a U-shaped binary image was used to simulate myocardial contraction. High-frequency components were obtained using wavelet analysis and then incorporated into the adaptive optical flow regularization process via function mapping. The final result was calculated, and the experimental results are as follows. Figure 5 As shown in the figure, compared with the previous ventricular wall motion estimation algorithm, the improved algorithm can extract the motion boundary of the myocardial ventricular wall and surrounding tissues more clearly, achieving the expected goal and proving the effectiveness of the algorithm.

[0067] The accuracy of the velocity vector field calculated by simulating myocardial ventricular wall motion is a key indicator of the quality of motion assessment algorithms. Root-Mean Square Error (RMSE) and Angular Error (AE) between optical flow vectors can be used as evaluation metrics for velocity vector field accuracy. RMSE reflects the degree of dispersion between the calculated velocity vector field and the actual vector field. AE reflects the angular error between the calculated optical flow vector and the actual optical flow vector. Using a U-shaped binary image, warp correction technology is employed to simulate myocardial contraction motion. Specifically, the left side of the ventricular wall is simulated to the right by 2 pixels, the right side by 2 pixels to the left, and the entire ventricular wall by 1 pixel downwards. The calculation range is taken from the region of interest of the ventricular wall. The accuracy of the wall motion estimation method of this embodiment is compared with that of the HS optical flow method, LK optical flow method, and Brox optical flow method to demonstrate the accuracy of the wall motion estimation method of this embodiment. Specific comparison results are shown in Table 1.

[0068] Table 1. Experimental evaluation results of the embodiments of the present invention, the HS optical flow method, the LK optical flow method, and the Brox optical flow method.

[0069]

[0070] The results show that the wall motion estimation method of the present invention has lower root mean square error and angle error values ​​compared with the other three optical flow methods, indicating that the wall motion estimation method of the present invention has higher calculation accuracy.

[0071] A method and apparatus for estimating ventricular wall motion based on an adaptive regularized optical flow model are proposed. Addressing the problem that commonly used optical flow model estimation methods employ constant regularization smoothing coefficients, leading to large calculation errors in the ventricular wall boundary region and unclear motion boundaries, this invention proposes using wavelet analysis to construct an adaptive optical flow regularized model for ventricular wall motion estimation. First, the initially obtained optical flow vector is analyzed using two-dimensional wavelet analysis and function mapping to obtain motion boundary information with relatively abrupt changes in the motion field. Second, this abrupt motion boundary information is adaptively fed back into the optical flow calculation model to correct the regularization coefficient of the smoothing term and recalculate the optical flow field. Finally, the above optical flow calculation process is repeated until the high-frequency wavelet components of the optical flow field converge. This method reflects the gradient difference between the myocardial ventricular wall and surrounding tissues in the regularization coefficient, thereby improving the accuracy of motion estimation.

[0072] Figure 6 This is a schematic diagram of a wall motion estimation device based on an adaptive regularized optical flow model, provided as an exemplary embodiment of this application. The wall motion estimation device based on an adaptive regularized optical flow model provided in this embodiment can execute the processing flow provided in an embodiment of a wall motion estimation method based on an adaptive regularized optical flow model. Figure 6 As shown, the wall motion estimation device 20 based on an adaptive regularized optical flow model provided in this application includes:

[0073] Image processing module 201 is used to obtain a sorted combination of multiple frames of ultrasound images within a preset cardiac cycle based on ultrasound cardiac images;

[0074] The first calculation module 202 is used to calculate the motion optical flow field of the two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method.

[0075] The decomposition and reconstruction module 203 is used to decompose and reconstruct the motion optical flow field of two consecutive frames of images using two-dimensional wavelet analysis and extract motion boundary information.

[0076] The mapping module 204 is used to map the motion boundary information into the regularization coefficient of the optical flow model and calculate the motion boundary smoothing coefficient.

[0077] The second calculation module 205 is used to recalculate the motion optical flow field of the two frames of images based on the motion boundary smoothing coefficient.

[0078] The processing module 206 repeatedly executes the decomposition and reconstruction module 203, the mapping module 204, and the second calculation module 205 until the wavelet coefficients converge, thus obtaining the final estimation result of the wall motion displacement.

[0079] The apparatus provided in this application embodiment can be specifically used to perform the above-described... Figure 1 The specific functions and technical effects of the solutions provided in the corresponding method embodiments will not be elaborated here.

[0080] This invention also provides a non-volatile storage device comprising: a processor, and a memory communicatively connected to the processor;

[0081] The memory stores instructions that the computer executes;

[0082] The processor executes computer execution instructions stored in the memory to implement the solution provided in any of the above method embodiments; the specific functions and technical effects achieved are not elaborated here. The electronic device can be the server mentioned above.

[0083] This application also provides a computer-readable storage medium storing computer-executable instructions. When executed by a processor, the computer-executable instructions are used to implement the solution provided in any of the above method embodiments. The specific functions and technical effects to be achieved are not described here.

[0084] This application also provides a computer program product, which includes a computer program stored in a readable storage medium. At least one processor of the electronic device can read the computer program from the readable storage medium. The at least one processor executes the computer program to cause the electronic device to perform the solution provided in any of the above method embodiments. The specific functions and technical effects that can be achieved are not described here.

[0085] The application scenarios described in this application are for the purpose of more clearly illustrating the technical solutions of this application, and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the emergence of new application scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0086] Those skilled in the art will understand that various aspects of this application can be implemented as a system, method, or program product. Therefore, various aspects of this application can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, collectively referred to herein as a "circuit," "module," or "system."

[0087] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0088] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for estimating wall motion based on an adaptive regularized optical flow model, characterized in that, The method includes: Step 101: Obtain a sorted combination of multiple ultrasound images within a preset cardiac cycle based on the ultrasound echocardiogram. Step 102: Calculate the motion optical flow field of the two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method; Step 103: Use two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of the two consecutive frames and extract motion boundary information; Step 104: The motion boundary information is mapped and added to the regularization coefficient of the optical flow model to calculate the motion boundary smoothing coefficient; Step 105: Recalculate the motion optical flow field of the two frames based on the motion boundary smoothing coefficient; Step 106: Repeat steps 103, 104 and 105 until the wavelet coefficients converge to obtain the final estimate of the wall motion displacement.

2. The method according to claim 1, characterized in that, The calculation of the motion optical flow field between two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method includes: The motion optical flow field of the two consecutive frames of ultrasound images is obtained by minimizing the total energy function of the two consecutive frames in the sorted combination of the multi-frame ultrasound images. The total energy function of the two consecutive frames is: E gradient and E color Adding them together represents the data items, E smooth This represents the smoothing term, where α is a regularization smoothing coefficient that is greater than zero.

3. The method according to claim 2, characterized in that, Data item E color According to the formula Calculate, data item E gradient According to the formula Calculate the smoothing term E. smooth According to the formula Calculate, where, ε is a constant, x represents the position of pixel A, and w represents the displacement of pixel A. This indicates the calculation of the derivative in the XY direction.

4. The method according to claim 2, characterized in that, The regularization smoothing coefficient α is expressed by the formula Calculate, where α g For global information weights, α l For local information weights, I x I y This represents the image gradient.

5. The method according to claim 2, characterized in that, The step of using two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of two consecutive frames of images and extract motion boundary information includes: According to formula f j =f j-1 +cmpt uses two-dimensional wavelet analysis to decompose and reconstruct the motion optical flow field of two consecutive frames and extract motion boundary information, where f j The highest approximation coefficients of the original signal f are derived from f. j It begins to decompose into lower-level approximations f. j-1 And wavelet component cmpt.

6. The method according to claim 5, characterized in that, By f j It begins to decompose into lower-level approximations f. j-1 The wavelet component cmpt1 is used, and this decomposition process is repeated until the four wavelet components cmpt1, cmpt2, cmpt3, and cmpt4 are extracted. Then, the extracted high-frequency component cmpt4 is used to obtain the non-boundary smoothing coefficient Wcmpt in the optical flow estimation through function mapping, according to the formula... +W cmpt The motion boundary information is mapped and added to the regularization coefficient of the optical flow model to calculate the motion boundary smoothness coefficient.

7. The method according to claim 6, characterized in that, The steps 103, 104, and 105 are executed iteratively until the wavelet coefficients converge, yielding the final estimation result of the ventricular wall motion displacement, including: Repeat steps 103, 104, and 105 until the difference between two adjacent wavelet high-frequency components converges, and obtain the final estimate of the ventricular wall motion displacement.

8. A device for estimating wall motion based on an adaptive regularized optical flow model, characterized in that, The wall motion estimation device includes: The image processing module is used to obtain a sorted combination of multiple ultrasound images within a preset cardiac cycle based on the ultrasound cardiac images; The first calculation module is used to calculate the motion optical flow field of the two consecutive frames of the sorted and combined multi-frame ultrasound images using the optical flow method. The decomposition and reconstruction module is used to decompose and reconstruct the motion optical flow field of two consecutive frames of images using two-dimensional wavelet analysis and extract motion boundary information. The mapping module is used to map the motion boundary information into the regularization coefficient of the optical flow model and calculate the motion boundary smoothing coefficient. The second calculation module is used to recalculate the motion optical flow field of the two frames of images based on the motion boundary smoothing coefficient. The processing module iteratively executes the decomposition and reconstruction module, the mapping module, and the second calculation module until the wavelet coefficients converge, thus obtaining the final estimation result of the wall motion displacement.

9. A non-volatile storage device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

Images