Method for estimating motion of a chamber wall segment based on confidence weight analysis of optical flow tracking

Abstract

The application relates to a chamber wall segment motion estimation method based on confidence weight analysis of optical flow tracking, belongs to the technical field of ultrasonic image processing, and aims to solve problems such as the need to rely on subjective judgment of doctors and quality of echocardiographic images for myocardial functional abnormalities. Through optical flow calculation based on confidence weight analysis, motion vector decomposition and ROI region dynamic tracking, the dynamic motion field of each segment is accurately calculated, and finally, the myocardial segment motion time curve is obtained. The optical flow calculation method based on confidence weight analysis corrects the motion amount from top to bottom through a Gaussian pyramid algorithm, combines image correction technology, adds an image "confidence weight matrix", greatly reduces the calculation proportion of the "shadow" place by using the confidence weight coefficient, and thus improves the accuracy of optical flow calculation. Through motion vector decomposition, ROI region dynamic tracking is realized, the motion characteristic information of the myocardial segment is more accurately reflected, and the calculation precision of the myocardial chamber wall motion is improved.

Description

Technical Field

[0001] This application relates to the field of ultrasonic image processing technology, and in particular to a method for estimating segmental motion of the ventricular wall based on confidence weight analysis and optical flow tracing. Background Technology

[0002] Cardiovascular diseases are a major threat to human health, ranking first among causes of death worldwide. Ischemic heart disease accounts for the largest proportion of cardiovascular diseases, primarily caused by insufficient blood and oxygen supply to the myocardium, leading to myocardial ischemia. Abnormal myocardial function seriously endangers human health, potentially causing heart failure, myocardial infarction, and even sudden death. Therefore, effectively assessing and analyzing myocardial function in real time is crucial. In recent years, echocardiography, with its advantages of being non-invasive, safe, inexpensive, and simple, has been widely used in the diagnosis of abnormal myocardial function and the evaluation of viable myocardium. The echocardiography procedure is simple, radiation-free, and inexpensive, and it can accurately obtain data on cardiac structure and myocardial function, accurately assessing myocardial contractile function and ventricular wall motion. Therefore, echocardiography has high value in the diagnosis of cardiovascular diseases and has become one of the important auxiliary diagnostic tools for physicians.

[0003] Myocardial function abnormalities can be evaluated using echocardiography to assess the coordination of myocardial ventricular wall motion. Commonly used techniques for assessing myocardial wall motion include tissue Doppler imaging (DTI) and two-dimensional speckle tracking imaging (2D-STI). DTI is an echocardiographic technique that can quantitatively measure local myocardial motion velocity and assess the myocardial motion status throughout the entire cardiac cycle. Two-dimensional speckle tracking imaging (2D-STI) is a quantitative assessment method that tracks the relative position and motion velocity of targets of interest on a two-dimensional ultrasound image. It uses methods such as M-mode ultrasound, spectral Doppler, tissue Doppler, speckle tracking, and three-dimensional ultrasound to measure the time differences in the peak systolic indices of different segments of the left ventricle to assess the coordination of myocardial function. However, DTI is often affected by cardiac rotation, oscillation, and respiratory movements, which can influence the image quality. 2D-STI also has limitations; it requires high image quality and high temporal and spatial resolution. If it cannot accurately reflect the motion information of all myocardial segments within the same cardiac cycle, tracking may fail.

[0004] To more effectively assess regional left ventricular myocardial systolic function, segmental motion analysis is commonly used clinically to provide more accurate judgments. For example, the American Society of Echocardiography recommends using a 17-segment model for analysis to reflect coronary blood supply. However, since the 17th segment (the apical cap) is difficult to detect under normal conditions, a 16-segment model is often used in echocardiographic analysis to assess left ventricular wall motion, evaluating myocardial function through strain analysis between different segments. Optical flow is a suitable choice for achieving real-time, high-resolution segmental motion estimation. However, obtaining accurate dynamic optical flow fields based on optical flow-based segmental motion estimation faces many challenges. For example, due to the rapid contraction of the heart, the motion amplitude of the target between frames is too large, and ordinary optical flow methods based on pixel-level calculations are often unsuitable for calculating large displacement fields. A larger problem is that echocardiographic image quality is often poor, and the acquisition process is inevitably affected by the patient's position, producing a certain degree of "shadowing." The lack of information in these "shadows" causes calculation errors, and blurred echocardiograms also lead to significant estimation errors. Summary of the Invention

[0005] This application provides a method for estimating segmental wall motion based on confidence weight analysis and optical flow tracing. It effectively solves the optical flow calculation errors caused by image quality problems such as large myocardial motion amplitude and the presence of "shadows". It can improve image quality and obtain more accurate ventricular wall motion mechanical parameters, so as to more effectively evaluate myocardial function.

[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 segmental wall motion based on confidence weight analysis optical flow tracing is provided, comprising:

[0008] Obtain a sequence of ultrasound images within a cardiac cycle based on echocardiograms;

[0009] The myocardial contour was obtained by segmenting the ultrasound image sequence and myocardial segments of different sections were marked using a 16-segment model;

[0010] The optical flow method based on confidence weight analysis was used to calculate the myocardial ventricular wall motion displacement distribution data and the optical flow field of the two frames before and after the segmented myocardial contour, and the motion field of the entire image sequence was calculated.

[0011] Based on the obtained motion field, the ROI region of each segment is dynamically tracked, and the motion vector of the myocardial motion within the ROI region is orthogonally decomposed to obtain the circumferential vector parallel to the inner boundary of the myocardium and the normal vector perpendicular to it.

[0012] The average displacement and motion curve of each segment are calculated to reflect the motion information of the myocardial segments.

[0013] Optionally, the calculation of the myocardial ventricular wall motion displacement distribution data of the segmented myocardial contour and the optical flow fields of the two frames before and after using the optical flow method based on confidence weight analysis includes:

[0014] Step 1: Select the initial resolution, determine the rule for the image resolution "pyramid" to gradually increase to the original resolution, and calculate the image sequence of different resolutions for each layer;

[0015] Step 2: Determine the confidence weights based on the input echocardiogram images. The grayscale distribution of the ultrasound images of each layer is normalized and then directly used as the confidence weights of the corresponding layers.

[0016] Step 3: Select the image of the current layer to calculate the optical flow field, take the optical flow field of the current layer as the incremental optical flow, multiply it with the confidence weight of the corresponding layer, and then add it with the result of upsampling the optical flow field of the previous layer to obtain the final output optical flow field of this layer;

[0017] Step 4: Repeat step 3 until the optical flow field at the original resolution is calculated iteratively, and use it as the final optical flow field output to calculate the myocardial ventricular wall motion displacement distribution data of the segmented myocardial contour.

[0018] Optionally, the optical flow method based on confidence weight analysis is performed according to... By performing optical flow pyramid iterations, the final output optical flow field of the corresponding layer is obtained, where u i v i u represents the final output optical flow field of the current layer in the pyramid model. i-1 v i-1 For the optical flow field output from the upper layer in the pyramid model, s i The confidence weights, du, are derived from the grayscale distribution matrix of the current layer image in the pyramid model. i dv i This represents the incremental optical flow obtained through the current layer image in the pyramid model.

[0019] Optionally, the confidence weights s i It is a confidence weight coefficient matrix formed by median filtering and normalization of the ultrasonic grayscale distribution map.

[0020] Optionally, the step of performing orthogonal decomposition of the myocardial motion within the ROI region to obtain a circumferential vector parallel to the inner boundary of the myocardium and a normal vector perpendicular to it includes:

[0021] At the myocardial boundary of each myocardial segment, motion is decomposed into a periodic vector parallel to it and a normal vector perpendicular to it, wherein the normal vector points in the direction of myocardial contraction. The normal vector is obtained by rotating the tangent vector by 90°, and the tangent vector is obtained using the formula... Calculate t i Let v be the i-th tangent vector. i is the i-th tangent point on the myocardial segment.

[0022] Optionally, the dynamic tracking of the ROI region of each segment based on the obtained motion field includes:

[0023] The labeled myocardial segments are then subjected to secondary segmentation and vector labeling;

[0024] Normalization is performed on the myocardial segments after vector labeling to obtain multiple circumferential and radial normalized vectors;

[0025] The normalized vector is combined with the region of interest to obtain the ROI vector field within the region, and the ROI region of the myocardial segment is dynamically tracked based on the ROI vector field.

[0026] Optionally, the step of dynamically tracking the ROI region of the myocardial segment based on the ROI vector field includes:

[0027] The ROI region of the myocardial segment is dynamically tracked according to the formula I' = warp(I, w), where I' is the new ROI, I is the old ROI, and w is the velocity vector calculated by comparing two consecutive frames of images.

[0028] Optionally, the segmented myocardial contour is marked using the 16-segment method to obtain 6 myocardial segments, which are then further segmented and labeled with vectors to obtain normalized motion vector information of 12 myocardial segments in the circumferential and radial directions.

[0029] Optionally, the root mean square error and the angular error between the optical flow vectors can be used as evaluation indicators for the accuracy of the velocity vector field to measure the quality of motion estimation.

[0030] According to a second aspect of the embodiments of this application, a device for estimating segmental wall motion based on confidence-weighted optical flow tracing is provided, comprising:

[0031] The acquisition module is used to acquire a sequence of ultrasound images within a cardiac cycle based on the ultrasound echocardiogram.

[0032] The segmentation module is used to segment the myocardial contour based on the ultrasound image sequence and mark the myocardial segments of different sections using a 16-segment model;

[0033] The calculation module is used to calculate the myocardial ventricular wall motion displacement distribution data and the optical flow field of the two frames before and after the segmented myocardial contour using the optical flow method based on confidence weight analysis, and to calculate the motion field of the entire image sequence.

[0034] The decomposition module is used to dynamically track the ROI region of each segment based on the obtained motion field, and to perform orthogonal decomposition of the motion vector of the myocardial motion within the ROI region to obtain the circumferential vector parallel to the inner boundary of the myocardium and the normal vector perpendicular to it.

[0035] The processing module is used to calculate the average displacement and motion curve of each segment, which reflects the motion information of the myocardial segments.

[0036] 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;

[0037] The memory stores computer-executed instructions;

[0038] The processor executes computer execution instructions stored in the memory to implement the method provided in the first aspect.

[0039] 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.

[0040] 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.

[0041] Beneficial effects:

[0042] This application discloses a method and device for estimating segmental wall motion based on confidence weight analysis and optical flow tracking. Addressing the issues of relying on physician subjective judgment and echocardiogram image quality for myocardial dysfunction, this method accurately calculates the dynamic motion field of each segment through confidence weight analysis-based optical flow calculation, motion vector decomposition, and dynamic tracking of the region of interest (ROI), ultimately obtaining the myocardial segment motion time curve. The confidence weight analysis-based optical flow calculation method uses a Gaussian pyramid algorithm to correct motion from top to bottom, combined with image correction techniques, and incorporates an image "confidence weight matrix." The confidence weight coefficients significantly reduce the computational weight of "shaded" areas, thereby improving the accuracy of optical flow calculation. Furthermore, motion vector decomposition enables dynamic tracking of the ROI, more accurately reflecting the motion characteristics of myocardial segments. This effectively tracks the motion state of each myocardial segment, significantly reducing computational errors caused by missing information at high resolution, thus improving the accuracy of myocardial ventricular wall motion calculation.

[0043] 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

[0044] 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.

[0045] Figure 1 This is a flowchart illustrating an exemplary embodiment of a method for estimating segmental wall motion based on confidence weight analysis optical flow tracing provided by this application.

[0046] Figure 2 This is a schematic diagram illustrating a confidence-optimized multi-resolution optical flow calculation according to an exemplary embodiment;

[0047] Figure 3 This is a schematic diagram of an ultrasound image “shadow” according to an exemplary embodiment;

[0048] Figure 4 This is a grayscale matrix image illustrated according to an exemplary embodiment;

[0049] Figure 5 This is a schematic diagram of the vector decomposition of myocardial ventricular wall motion according to an exemplary embodiment;

[0050] Figure 6 This is a schematic diagram of an ROI vector field according to an exemplary embodiment;

[0051] Figure 7 This is a schematic diagram illustrating the error in the calculation result of segmental displacement according to an exemplary embodiment;

[0052] Figure 8 This is a hypersonic simulation diagram of the corresponding "shade" shown according to an exemplary embodiment;

[0053] Figure 9 This is a schematic diagram illustrating the calculation results before and after using a grayscale matrix, according to an exemplary embodiment.

[0054] Figure 10 This is a schematic diagram illustrating dynamic ROI tracking according to an exemplary embodiment;

[0055] Figure 11 This is a simulation diagram of a myocardial wall with "shading" shown according to an exemplary embodiment;

[0056] Figure 12 This is a schematic diagram illustrating the comparison of experimental results of the dual-flow method according to an exemplary embodiment;

[0057] Figure 13 This is a comparative schematic diagram showing the normalized experimental results according to an exemplary embodiment;

[0058] Figure 14 This is a schematic diagram of the region of interest after segmentation of the myocardial ventricular wall in a normal patient, according to an exemplary embodiment.

[0059] Figure 15 This is a schematic diagram illustrating the comparison of experimental results according to an exemplary embodiment;

[0060] Figure 16 This is a schematic diagram illustrating a normalized comparison of another experimental result according to an exemplary embodiment;

[0061] Figure 17 This is a schematic diagram of a myocardial motion displacement curve according to an exemplary embodiment.

[0062] Figure 18 This is a schematic diagram of a wall segment motion estimation device based on confidence weighted optical flow tracing, according to an exemplary embodiment. Detailed Implementation

[0063] 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.

[0064] 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.

[0065] To improve image quality and obtain more accurate wall motion biomechanical parameters for more effective assessment of myocardial function, this application proposes an optical flow method based on confidence weight analysis, using echocardiographic images from 15 patients with myocardial motion abnormalities and 15 normal patients. Simultaneously, dynamic tracking of the Region of Interest (ROI) is achieved through motion vector decomposition, ultimately calculating the displacement of each segment to accurately reflect the motion information of myocardial segments. Experimental results show that this method is highly accurate and robust, image quality can be corrected, myocardial function coordination analysis is easily implemented, and it has suitable quantification capabilities for myocardial wall motion, which can assist in rapid clinical diagnosis.

[0066] Figure 1 A flowchart illustrating a method for estimating segmental wall motion based on confidence weight analysis and optical flow tracing, provided as an exemplary embodiment of this application. Figure 1 As shown, the specific steps of this method for calculating segmental motion of the myocardial ventricular wall are as follows:

[0067] Step 110: Obtain an ultrasound image sequence within one cardiac cycle based on the ultrasound echocardiogram.

[0068] In this embodiment of the invention, echocardiogram images of 30 patients were collected as raw image data. An ultrasound image sequence within one cardiac cycle was extracted and analyzed to calculate the displacement data of myocardial ventricular wall motion, which has very important reference value for the diagnosis of myocardial ischemia in cardiovascular diseases.

[0069] Step 120: The myocardial contour is obtained by segmenting the ultrasound image sequence and the myocardial segments of different sections are marked using a 16-segment model.

[0070] For example, you can first manually segment the myocardial outline, and then label the myocardial segments in different sections according to the 16-segment model recommended by the American Society of Echocardiography.

[0071] Step 130: The optical flow method based on confidence weight analysis is used to calculate the myocardial ventricular wall motion displacement distribution data of the segmented myocardial contour and the optical flow field of the two frames before and after, and the motion field of the entire image sequence is calculated.

[0072] Traditional optical flow methods are motion representations of image surface brightness patterns. They establish the relationship between image pixel motion and grayscale through optical flow constraint equations and the assumption of constant brightness.

[0073] The relationship between image pixel motion and grayscale is established using optical flow constraint equations and the assumption of constant brightness. This relationship can be expressed by the formula I(x+dx,y+dy,t+dt)=I(x,y,t), where (x,y) are the image pixel positions, (dx,dy) are the displacement vectors per unit time, t represents time, and dt represents the time interval between two consecutive images.

[0074] Brox et al. proposed the Brox optical flow method in 2004, which further optimized and improved the optical flow calculation method. Based on the Brox optical flow method, this invention introduces an adaptive multi-resolution analysis method to further improve the calculation accuracy of image pixel motion.

[0075] Echocardiography reveals large amplitude myocardial motion and numerous artifacts, leading to significant errors in traditional optical flow calculations for large displacements and shaded regions. To improve the accuracy of motion estimation, reference... Figure 2As shown, this embodiment of the invention constructs a multi-resolution "pyramid" strategy based on confidence optimization: two adjacent frames of images are calculated in layers according to different resolutions, with large displacements calculated at low resolutions and small displacements calculated at high resolutions. Starting from the top layer with the lowest resolution, the optical flow calculation results of the previous layer are fed back to the next layer as the incremental optical flow (corrected motion) of the next layer's image, and the calculation is repeated until the bottom layer of the pyramid is reached. During the inter-layer transfer of optical flow, image correction technology and confidence optimization screening strategy are introduced, that is, the obtained motion field is used to inversely deform the second frame image to generate a new image that is closer to the first frame, and the new image and the first frame image are used as an image pair to be transferred to a layer for incremental optical flow calculation. At the same time, the "confidence weight matrix" is multiplied with the incremental optical flow field of this layer and then transferred to the next layer. The steps for calculating the myocardial ventricular wall motion displacement distribution data and the optical flow fields of the two frames before and after segmentation of the myocardial contour using the optical flow method based on confidence weight analysis are as follows:

[0076] Step 1: Select the initial resolution, determine the rules for the image resolution "pyramid" to gradually increase to the original resolution, and calculate the image sequence at each different resolution level. An exponential function is used for the pyramid design.

[0077] Step 2: Determine the confidence weights based on the input echocardiogram images. The grayscale distribution of the ultrasound image for each layer is normalized and then directly used as the confidence weight for the corresponding layer.

[0078] Step 3: Select the image of the current layer to calculate the optical flow field. Multiply the optical flow field of the current layer as the incremental optical flow and the confidence weight of the corresponding layer, and then add it to the result of upsampling the optical flow field of the previous layer to obtain the final output optical flow field of this layer.

[0079] Step 4: Repeat step 3 until the optical flow field at the original resolution (bottom of the pyramid) is calculated. This is used as the final optical flow field output. Then, the myocardial ventricular wall motion displacement distribution data of the segmented myocardial contour and the optical flow fields of the previous and next frames are calculated.

[0080] By introducing the Gaussian pyramid algorithm, the optical flow calculation results of the top layer are fed back to the next layer as the optical flow estimate of the next layer. In this way, the motion is corrected from top to bottom along the Gaussian pyramid until the bottom layer of the pyramid is reached.

[0081] To implement the multi-layer pyramid strategy, this invention innovatively employs image correction technology. Specifically, after optical flow calculation is performed on the original echocardiogram image data to obtain the velocity vector w of each pixel, a new image is generated through inverse image deformation and compared with the original image to evaluate the algorithm's accuracy. The image correction technology used in this invention significantly improves the efficiency of optical flow calculation and can also track rapidly moving image pixels.

[0082] During echocardiography, clinicians typically acquire ultrasound images with the patient in the left lateral decubitus position. However, rib obstruction often results in "shadows" in the echocardiogram data. Figure 3 As shown. Typically, the amount of information contained in this "shadow" region is limited. Therefore, this embodiment of the invention proposes an adaptive weighting coefficient based on image grayscale to minimize the impact of the "shadow" on optical flow calculation.

[0083] Like color images, grayscale images describe the overall and local distribution and characteristics of chromaticity and brightness levels. As defined by optical flow, the instantaneous rate of change of grayscale at a specific coordinate point on a two-dimensional image plane is typically defined as the optical flow vector. The grayscale value directly reflects the image quality. Due to "shadow" occlusion, the grayscale change between two frames is relatively small, thus affecting the accuracy of optical flow calculation. Correspondingly, the grayscale value also reflects image information to some extent; the image information at "shadow" locations is relatively low. This invention innovatively uses a "confidence weight matrix" based on a "grayscale matrix" to solve this problem, i.e., a matrix obtained after normalizing the grayscale image, such as... Figure 4 As shown.

[0084] The value of each pixel in the grayscale matrix represents its weight in the next pyramid iteration. During the iteration process, the grayscale matrix at the corresponding resolution is updated synchronously. The optical flow method based on confidence weight analysis follows... By performing optical flow pyramid iterations, the final output optical flow field of the corresponding layer is obtained, where u i v i u represents the final output optical flow field of the current layer in the pyramid model. i-1 v i-1 For the optical flow field output from the upper layer in the pyramid model, s i The confidence weights, du, are derived from the grayscale distribution matrix of the current layer image in the pyramid model. i dv i This represents the incremental optical flow obtained through the current layer image in the pyramid model. The confidence weight s i It is a confidence weight coefficient matrix formed by median filtering and normalization of the ultrasonic grayscale distribution map.

[0085] Selecting an echocardiogram with localized "shading," and cropping a frame from it, along with its corresponding simulated image, reveals that at low resolution, the size of the "shading" area is correspondingly reduced, while in the high-resolution image, the motion in the corresponding "shading" area better reflects the overall myocardial motion within that region. At high resolution, the calculation of optical flow is more accurate in areas with noticeable brightness, while the "shading" area suffers from information loss, leading to errors in optical flow calculation. The simulated image is shown below. Figure 8 As shown, Figure 8 (a), (b), and (c) represent 30%, 60%, and 100% of the original image resolution, respectively.

[0086] Select an echocardiogram with localized "shading," crop a frame from it, and apply the "grayscale matrix" of this invention to compare it with the adaptive optical flow calculation result based on image grayscale, as shown in the example. Figure 9 As shown, Figure 9 (a) shows the calculation results without using the grayscale matrix. Figure 9 (b) shows the calculation results after using the grayscale matrix. (Reference) Figure 9 As shown, by using the method of adding the gray matrix, the weight of the "shaded" area is relatively small during the iteration process, which can reduce the calculation errors caused by missing information at high resolution and effectively solve the "shaded" problem.

[0087] Optical flow filtering based on confidence optimization coefficients is key to this invention. Similar to color images, grayscale images reflect the overall and local distribution and characteristics of chromaticity and brightness levels. As defined by optical flow, the instantaneous rate of change of grayscale at a specific coordinate point on a two-dimensional image plane is defined as the optical flow vector. The grayscale value directly reflects the image quality. Due to "shadow" occlusion, the grayscale change between two frames is relatively small, thus affecting the accuracy of optical flow calculation. Correspondingly, the grayscale value also reflects image information to some extent; the image information at "shadow" locations is relatively low. This invention uses a median-filtered and normalized ultrasound grayscale distribution map as a confidence optimization matrix. This matrix is ​​weighted with the optical flow field of the current layer and used as the final optical flow field for the next layer, thereby improving the accuracy of optical flow calculation. This allows for better tracking of the motion state of different myocardial segments, effectively reducing calculation errors caused by missing information at high resolution, and thus improving the accuracy of myocardial ventricular wall motion calculation.

[0088] Step 140: Based on the obtained motion field, dynamically track the ROI region of each segment, and perform orthogonal decomposition of the motion vector of the myocardium within the ROI region to obtain the circumferential vector parallel to the inner boundary of the myocardium and the normal vector perpendicular to it.

[0089] Furthermore, at the myocardial boundary of each myocardial segment, the motion is decomposed into a periodic vector parallel to it and a normal vector perpendicular to it, wherein the normal vector points in the direction of myocardial contraction. The normal vector is obtained by rotating the tangent vector by 90°, and the tangent vector is obtained using the formula... Calculate t i Let v be the i-th tangent vector. i is the i-th tangent point on the myocardial segment.

[0090] Furthermore, the labeled myocardial segments are further segmented and vector-labeled; the vector-labeled myocardial segments are normalized to obtain multiple circumferential and radial normalized vectors;

[0091] The normalized vector is combined with the region of interest to obtain the ROI vector field within the region, and the ROI region of the myocardial segment is dynamically tracked according to the ROI vector field. The ROI region of the myocardial segment is dynamically tracked according to the formula I' = warp(I, w), where I' is the new region of interest, I is the old region of interest, and w is the velocity vector calculated by comparing the two frames of images.

[0092] The ventricular wall of the myocardium is arched in shape. Under the influence of pumping and other processes, the wall moves, and its displacement can be decomposed into two mutually perpendicular components: radial and circumferential. To better analyze the myocardial motion of different segments of the ventricular wall, this invention proposes a vector separation method, with the radial direction as the primary analysis target, such as... Figure 5 As shown.

[0093] After segmenting the myocardium, assuming the circumferential motion of each segment is along the ventricular wall, the corresponding normal vector points in the direction of myocardial contraction. The normal vector can be obtained by rotating the tangent vector by 90°, using the formula... Calculate, where t i Let v be the i-th tangent vector. i is the i-th tangent point on the myocardial segment.

[0094] The segmented myocardial contour was labeled using a 16-segment method, resulting in 6 myocardial segments. These 6 segments were then further segmented and vector-labeled to obtain 12 normalized circumferential and radial motion vectors, allowing for better observation of intra-segment motion. The 12 segments obtained from the secondary segmentation were then vector-labeled and normalized to obtain 12 normalized circumferential and radial vectors. These vectors were then combined with the region of interest (ROI) to obtain the ROI vector field. The displacement data calculated using optical flow was subjected to directional separation to reduce the influence of anomalous motion caused by uneven stress on the ventricular wall. The final ROI vector field is shown below. Figure 6 As shown.

[0095] During contraction, the relative position and shape of the myocardial ventricular wall change significantly. If the displacement of each segment is calculated based on the initial state of the myocardial region, the actual myocardial position may differ from the region of interest at a certain contraction stage, leading to substantial errors in the calculated segmental displacement. Figure 7 As shown.

[0096] To better track the motion of different segments of the myocardium, this embodiment of the invention employs a region of interest (ROI) dynamic tracking algorithm for the myocardial ventricular wall, ensuring that the ROI region remains consistent with the actual ventricular wall region. A velocity vector w is calculated by comparing two frames of images. Then, image correction techniques are used to map each pixel in the initial ROI onto another plane under the influence of the velocity vector, thereby generating a new ROI I'. The ROI dynamic tracking algorithm can be represented by the formula I' = warp(I, w), where I' is the new ROI, I is the old ROI, and w is the velocity vector calculated by comparing the two frames of images.

[0097] This invention uses the root-mean-square error (RMSE) and the angular error between the optical flow vectors as evaluation metrics for the accuracy of the velocity vector field, to measure the quality of motion estimation. RMSE reflects the degree of dispersion between the estimated velocity vector field and the actual vector field, and its range is taken within the region of interest. The formula for calculating the error is:

[0098] To assess ventricular wall motion, each segment should be evaluated over multiple cardiac cycles, and a scoring system for wall motion status should be used for quantitative assessment to compare segmental wall motion abnormalities among patients. A four-level scoring system is generally used: (1) normal or hyperactive, (2) hypoactive (reduced thickening), (3) absent (negligible thickening), and (4) dyskinesia (contractional thinning or stretching). Normal or hyperactive wall motion: 1 point. Hypoactive wall motion, i.e., endocardial motion amplitude less than 5 mm: 2 points. Absent wall motion, i.e., endocardial motion amplitude less than 2 mm: 3 points. Paradoxical wall motion, i.e., antagonistic motion: 4 points. The wall motion score index (WMSI) is calculated after scoring. The formula for calculating the WMSI is:

[0099]

[0100] Select a normal echocardiogram image, capture the process of the myocardium transitioning from diastole to systole within one cardiac cycle, and record the experimental results as follows: Figure 10 As shown. The ROI region dynamic tracking algorithm applied in this embodiment of the invention can effectively track the motion state of each segment of the myocardium, thereby improving the calculation accuracy of myocardial ventricular wall motion.

[0101] Step 150: Calculate the displacement data of each myocardial segment, and determine the displacement curve of the myocardial ventricular wall segment motion based on the displacement data.

[0102] By calculating the displacement data of each myocardial segment and determining the displacement curve of the myocardial ventricular wall segment motion based on the displacement data, the motion information of the myocardial segment is accurately reflected.

[0103] This invention embodiment uses a simulated image of the right ventricular wall with a "shadow" added to it for experimental purposes. Figure 11 As shown. This experiment used warp technology to simulate myocardial contraction. Specifically, the left wall shifted 2 pixels to the right, the right wall shifted 2 pixels to the left, and the entire ventricular wall shifted 1 pixel downward.

[0104] The experimental results of the optical flow method based on confidence weight analysis in this invention are compared with those of the HS optical flow method, LK optical flow method, and Brox optical flow method. Figure 12 As shown, the experimental results after normalization are as follows: Figure 13 As shown. Furthermore, this embodiment of the invention selects an echocardiogram of a normal patient for experimentation. The warp technique is also used to simulate myocardial ventricular wall motion: the left wall moves 1 pixel to the right, the right wall moves 1 pixel to the left, and the entire ventricular wall moves downwards by 0.5 pixels. The region of interest segmented within the error calculation range is shown below. Figure 14 As shown, the optical flow method based on confidence weight analysis in this embodiment of the invention is compared with the HS optical flow method, LK optical flow method, and Brox optical flow method. The experimental results are as follows: Figure 15 As shown, the experimental results after normalization are as follows: Figure 16 As shown.

[0105] Depend on Figure 12 and Figure 13 , Figure 15 and Figure 16 The experimental comparison results show that the HS optical flow method and LK optical flow method have high calculation error rates. Moreover, the HS optical flow method, LK optical flow method and Brox optical flow method all have directional errors in the calculation of the "shadow" area and the segmented region of interest. However, the improved optical flow method based on adaptive multi-resolution analysis in this embodiment of the invention has significantly better results than the other three algorithms. It can reduce the calculation errors caused by missing information at high resolution, effectively solve the problem of motion calculation errors caused by the darkness of the "shadow" area, and thus improve the accuracy of the calculation.

[0106] In this embodiment of the invention, the final displacement curve can be obtained by calculating the average displacement of pixels within the ROI region of each segment of the myocardium, such as... Figure 17 As shown. Figure 17 The 12 curves represent the displacement of each segment after directional separation. Besides reflecting the overall displacement of each myocardial segment, the curves also reveal the motion details within the segment. The results show that the displacement curves, plotted over two cardiac cycles, exhibit a regular arched shape, demonstrating the accuracy of the ventricular wall segment motion estimation algorithm based on confidence-weighted optical flow tracing in this embodiment of the invention.

[0107] This application discloses a method and device for estimating segmental wall motion based on confidence weight analysis and optical flow tracking. Addressing the issues of relying on physician subjective judgment and echocardiogram image quality for myocardial dysfunction, this method accurately calculates the dynamic motion field of each segment through confidence weight analysis-based optical flow calculation, motion vector decomposition, and dynamic tracking of the region of interest (ROI), ultimately obtaining the myocardial segment motion time curve. The confidence weight analysis-based optical flow calculation method uses a Gaussian pyramid algorithm to correct motion from top to bottom, combined with image correction techniques, and incorporates an image "confidence weight matrix." The confidence weight coefficients significantly reduce the computational weight of "shaded" areas, thereby improving the accuracy of optical flow calculation. Furthermore, motion vector decomposition enables dynamic tracking of the ROI, more accurately reflecting the motion characteristics of myocardial segments. This effectively tracks the motion state of each myocardial segment, significantly reducing computational errors caused by missing information at high resolution, thus improving the accuracy of myocardial ventricular wall motion calculation.

[0108] Figure 18 This is a schematic diagram of a ventricular wall segment motion estimation device based on confidence-weighted optical flow tracing, provided as an exemplary embodiment of this application. The ventricular wall segment motion estimation device based on confidence-weighted optical flow tracing provided in this embodiment can execute the processing flow provided in an embodiment of a ventricular wall segment motion estimation method based on confidence-weighted optical flow tracing. For example... Figure 18 As shown, the wall segment motion estimation device 20 based on confidence weighted optical flow tracing provided in this application includes:

[0109] The acquisition module 201 is used to acquire a sequence of ultrasound images within a cardiac cycle based on the ultrasound echocardiogram.

[0110] The segmentation module 202 is used to segment the myocardial contour according to the ultrasound image sequence and mark the myocardial segments of different sections using a 16-segment model;

[0111] The calculation module 203 is used to calculate the myocardial ventricular wall motion displacement distribution data and the optical flow field of the previous and next frames of the segmented myocardial contour using the optical flow method based on confidence weight analysis, and to calculate the motion field of the entire image sequence.

[0112] The decomposition module 204 is used to dynamically track the ROI region of each segment according to the obtained motion field, and to perform orthogonal decomposition of the motion vector of the myocardial motion in the ROI region to obtain the circumferential vector parallel to the inner boundary of the myocardium and the normal vector perpendicular to it.

[0113] The processing module 205 is used to calculate the average displacement and motion curve of each segment, which reflects the motion information of the myocardial segments.

[0114] 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.

[0115] This invention also provides a non-volatile storage device comprising: a processor, and a memory communicatively connected to the processor;

[0116] The memory stores instructions that the computer executes;

[0117] 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.

[0118] 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.

[0119] 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.

[0120] 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.

[0121] 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."

[0122] In some possible implementations, the electronic device according to this application may include at least one processor and at least one memory. The memory stores program code that, when executed by the processor, causes the processor to perform the operational data management methods according to the various exemplary embodiments of this application described above. For example, the processor may perform steps such as those in the operational data management method.

[0123] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0124] Furthermore, although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0125] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0126] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable image scaling device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable image scaling device, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0127] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable image scaling device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0128] These computer program instructions can also be loaded onto a computer or other programmable image scaling device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable device for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0129] 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.

[0130] 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 segmental wall motion based on confidence weight analysis and optical flow tracing, characterized in that, The method includes: Obtain a sequence of ultrasound images within a cardiac cycle based on echocardiograms; The myocardial contour was obtained by segmenting the ultrasound image sequence and myocardial segments of different sections were marked using a 16-segment model; The optical flow method based on confidence weight analysis was used to calculate the myocardial ventricular wall motion displacement distribution data and the optical flow field of the two frames before and after the segmented myocardial contour, and the motion field of the entire image sequence was calculated. Based on the obtained motion field, the ROI region of each segment is dynamically tracked, and the motion vector of the myocardial motion within the ROI region is orthogonally decomposed to obtain the circumferential vector parallel to the inner boundary of the myocardium and the normal vector perpendicular to it. The average displacement and motion curve of each segment are calculated to reflect the motion information of the myocardial segments; The myocardial ventricular wall motion displacement distribution data obtained by calculating the segmented myocardial contour using the optical flow method based on confidence weight analysis includes: Step 1: Select the initial resolution, determine the rule for the image resolution "pyramid" to gradually increase to the original resolution, and calculate the image sequence of different resolutions for each layer; Step 2: Determine the confidence weights based on the input echocardiogram images. The grayscale distribution of the ultrasound images of each layer is normalized and then directly used as the confidence weights of the corresponding layers. Step 3: Select the image of the current layer to calculate the optical flow field, take the optical flow field of the current layer as the incremental optical flow, multiply it with the confidence weight of the corresponding layer, and then add it with the result of upsampling the optical flow field of the previous layer to obtain the final output optical flow field of this layer; Step 4: Repeat step 3 until the optical flow field at the original resolution is calculated iteratively, and use it as the final optical flow field output to calculate the myocardial ventricular wall motion displacement distribution data of the segmented myocardial contour. The optical flow method based on confidence weight analysis is as follows By iterating through the optical flow pyramid, the final output optical flow field of the corresponding layer is obtained, where... , This represents the final output optical flow field of the current layer in the pyramid model. , This represents the optical flow field output from the upper layer of the pyramid model. The confidence weights are derived from the grayscale distribution matrix of the current layer image in the pyramid model. , This refers to the incremental optical flow obtained through the current layer image in the pyramid model; The confidence weight It is a confidence weight coefficient matrix formed by median filtering and normalization of the ultrasonic grayscale distribution map.

2. The method according to claim 1, characterized in that, The process of performing orthogonal decomposition of myocardial motion within the ROI region to obtain a circumferential vector parallel to the inner boundary of the myocardium and a normal vector perpendicular to it includes: At the myocardial boundary of each myocardial segment, motion is decomposed into a periodic vector parallel to it and a normal vector perpendicular to it, wherein the normal vector points in the direction of myocardial contraction. The normal vector is obtained by rotating the tangent vector by 90°, and the tangent vector is obtained using the formula... Calculate t i Let v be the i-th tangent vector. i is the i-th tangent point on the myocardial segment.

3. The method according to claim 2, characterized in that, The dynamic tracking of the ROI region for each segment based on the obtained sports field includes: The labeled myocardial segments are then subjected to secondary segmentation and vector labeling; Normalization is performed on the myocardial segments after vector labeling to obtain multiple circumferential and radial normalized vectors; The normalized vector is combined with the region of interest to obtain the ROI vector field within the region, and the ROI region of the myocardial segment is dynamically tracked based on the ROI vector field.

4. The method according to claim 3, characterized in that, The dynamic tracking of the ROI region of the myocardial segment based on the ROI vector field includes: According to the formula The ROI region of the myocardial segment is dynamically tracked, where I' is the new ROI, I is the old ROI, and w is the velocity vector calculated by comparing two consecutive frames of images.

5. The method according to claim 3, characterized in that, The segmented myocardial contour was labeled using a 16-segment method to obtain 6 myocardial segments. Then, a second segmentation and vector labeling were performed to obtain normalized motion vector information of 12 myocardial segments in the circumferential and radial directions.

6. The method according to claim 3, characterized in that, The root mean square error and the angular error between the optical flow vectors are used as evaluation indicators for the accuracy of the velocity vector field, which are used to measure the quality of motion estimation.

7. A device for estimating segmental wall motion based on confidence-weighted optical flow tracing, used to implement the method described in any one of claims 1-6, characterized in that, The ventricular wall segment motion estimation device includes: The acquisition module is used to acquire a sequence of ultrasound images within a cardiac cycle based on the ultrasound echocardiogram. The segmentation module is used to segment the myocardial contour based on the ultrasound image sequence and mark the myocardial segments of different sections using a 16-segment model; The calculation module is used to calculate the myocardial ventricular wall motion displacement distribution data and the optical flow field of the two frames before and after the segmented myocardial contour using the optical flow method based on confidence weight analysis, and to calculate the motion field of the entire image sequence. The decomposition module is used to dynamically track the ROI region of each segment based on the obtained motion field, and to perform orthogonal decomposition of the motion vector of the myocardial motion within the ROI region to obtain the circumferential vector parallel to the inner boundary of the myocardium and the normal vector perpendicular to it. The processing module is used to calculate the average displacement and motion curve of each segment, which reflects the motion information of the myocardial segments.

Images