Atrial defect and ventricular defect image model reconstruction method

Abstract

The invention provides an atrial defect and ventricular defect image model reconstruction method, and relates to the technical field of medical image processing. The method comprises the following steps: acquiring two-dimensional medical image data of a heart area; based on the two-dimensional medical image data, utilizing an image segmentation technology to identify atrial defect and ventricular defect areas and extracting feature points; constructing a preliminary three-dimensional model of the atrial defect and the ventricular defect according to the feature points by using a three-dimensional reconstruction algorithm; and the preliminary three-dimensional model is refined through a model optimization technology, so that the accuracy and definition of the preliminary three-dimensional model are improved. By means of the scheme, the accuracy and definition of the three-dimensional model can be improved in atrial defect and ventricular defect image model reconstruction.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and particularly to a method for reconstructing atrial and ventricular defect imaging models. Background Art

[0002] For the method of reconstructing atrial and ventricular defect imaging models, it generally refers to the process of obtaining two-dimensional or three-dimensional image data of the cardiac atrioventricular region through medical imaging technology, and using computer-aided design and image processing technology to construct a three-dimensional visualization model of atrial and ventricular defects of the heart. This method helps doctors more intuitively understand the abnormal conditions of the patient's heart structure, thus providing support for surgical planning. However, there is a key problem in practical applications: how to improve the accuracy and clarity of the three-dimensional model in the reconstruction of atrial and ventricular defect imaging models. This is because the cardiac structure is complex and dynamically changing, and coupled with the possible noise interference and technical limitations in the imaging process, these factors may all lead to deviations between the reconstructed three-dimensional model and the actual situation, thereby affecting the accuracy of clinical decisions. Therefore, improving the quality of the reconstructed model is an important research direction in this field. Summary of the Invention

[0003] In view of this, the present invention provides a method for reconstructing atrial and ventricular defect imaging models, which at least partially solves the problems existing in the prior art.

[0004] The method for reconstructing atrial and ventricular defect imaging models includes:

[0005] Obtain two-dimensional medical image data of the cardiac region;

[0006] Based on the two-dimensional medical image data, use image segmentation technology to identify atrial and ventricular defect regions and extract feature points;

[0007] Use a three-dimensional reconstruction algorithm to construct a preliminary three-dimensional model of atrial and ventricular defects according to the feature points;

[0008] Refine the preliminary three-dimensional model through model optimization technology to improve its accuracy and clarity.

[0009] In a specific embodiment, obtaining two-dimensional medical image data of the cardiac region includes:

[0010] Preprocess the original medical image data, including gray-scale correction and noise removal;

[0011] Based on the preprocessed medical image data, use a cardiac region localization algorithm to identify the cardiac contour;

[0012] Use the cardiac contour information to crop out the image segment containing the cardiac region;

[0013] Enhance the cropped image segments to highlight the atrial septal defect and ventricular septal defect regions.

[0014] In a specific embodiment, based on the two-dimensional medical image data, using image segmentation technology to identify the atrial septal defect and ventricular septal defect regions and extract feature points includes:

[0015] Apply threshold segmentation technology to initially identify the atrial septal defect and ventricular septal defect regions;

[0016] Based on the initial identification result, use an edge detection algorithm to further accurately locate the boundaries of the atrial septal defect and ventricular septal defect;

[0017] If the overlap rate between the edge detection result and the threshold segmentation result is less than a predetermined ratio, adjust the threshold and re-segment;

[0018] Extract key feature points from the finally determined atrial septal defect and ventricular septal defect regions.

[0019] In a specific embodiment, based on the initial identification result, using an edge detection algorithm to further accurately locate the boundaries of the atrial septal defect and ventricular septal defect includes:

[0020] Use a Gaussian filter to smooth the image to reduce the influence of noise;

[0021] Based on the smoothed image, apply the Sobel operator to calculate the gradient intensity;

[0022] If the gradient intensity is greater than the set threshold T, mark this pixel as a boundary point;

[0023] Connect the continuous boundary points to form the complete boundaries of the atrial septal defect and ventricular septal defect.

[0024] In a specific embodiment, if the gradient intensity is greater than the set threshold T, marking this pixel as a boundary point includes:

[0025] Calculate the gradient intensity G for each pixel point;

[0026] If G>T, mark this pixel as a boundary point;

[0027] Based on the marked boundary points, use connected component analysis to determine the boundary;

[0028] If the number of connected components is less than the expected number N, adjust the threshold T and re-mark.

[0029] In a specific embodiment, if the number of connected components is less than the expected number N, adjusting the threshold T and re-marking includes:

[0030] Calculate the number M of the current connected components;

[0031] If M < N, then lower the threshold T and return to recalculate the gradient intensity;

[0032] If M >= N, then keep the threshold T unchanged;

[0033] Refine the finally determined boundary points to obtain a more accurate boundary.

[0034] In a specific embodiment, the step of identifying the atrial septal defect and ventricular septal defect regions and extracting feature points based on the two-dimensional medical image data includes:

[0035] Preprocess the two-dimensional medical image data, including noise reduction and smoothing processing;

[0036] Based on the preprocessed image data, use the threshold segmentation method to preliminarily identify the atrial septal defect and ventricular septal defect regions;

[0037] Use the edge detection algorithm to further accurately locate the boundaries of the atrial septal defect and ventricular septal defect;

[0038] Select multiple key feature points on the located boundary for subsequent three-dimensional reconstruction.

[0039] In a specific embodiment, the step of using the threshold segmentation method to preliminarily identify the atrial septal defect and ventricular septal defect regions based on the preprocessed image data includes:

[0040] Calculate the gray value of each pixel in the image data;

[0041] Based on the gray histogram, determine a threshold T such that the pixels with gray values greater than T belong to the atrial septal defect and ventricular septal defect regions;

[0042] If the gray value G is greater than the threshold T, then mark this pixel as part of the atrial septal defect and ventricular septal defect regions;

[0043] Connect the marked pixels to form the contour of the preliminary atrial septal defect and ventricular septal defect regions.

[0044] In a specific embodiment, the step of if the gray value G is greater than the threshold T, then mark this pixel as part of the atrial septal defect and ventricular septal defect regions includes:

[0045] For each pixel i, calculate the difference Δi = Gi - T between its gray value Gi and the threshold T;

[0046] If Δi > 0, then mark pixel i as part of the atrial septal defect and ventricular septal defect regions;

[0047] If Δi ≤ 0, then do not mark pixel i;

[0048] Update the outlines of the atrial and ventricular defect regions based on the marking results.

[0049] In a specific embodiment, the step of calculating the difference Δi = Gi - T between the gray value Gi of each pixel i and the threshold T includes:

[0050] Obtain the gray value Gi of pixel i;

[0051] Set the threshold T;

[0052] Calculate the difference Δi = Gi - T;

[0053] If Δi > 0, proceed to the next step; otherwise, return to the unmarked state.

[0054] In a specific embodiment, the step of if Δi > 0, proceed to the next step; otherwise, return to the unmarked state includes:

[0055] If Δi > 0, mark pixel i as part of the atrial and ventricular defect regions;

[0056] If Δi ≤ 0, do not mark pixel i and check the next pixel.

[0057] The present invention provides a method for reconstructing an atrial and ventricular defect imaging model, including: obtaining two-dimensional medical image data of the heart region; based on the two-dimensional medical image data, using image segmentation technology to identify the atrial and ventricular defect regions and extract feature points; using a three-dimensional reconstruction algorithm to construct a preliminary three-dimensional model of the atrial and ventricular defects according to the feature points; and performing refinement processing on the preliminary three-dimensional model through model optimization technology to improve its accuracy and clarity. Through the solution of the present invention, it is possible to solve the problem of how to improve the accuracy and clarity of the three-dimensional model in the reconstruction of the atrial and ventricular defect imaging model. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In the drawings, unless otherwise specified, the same reference numerals throughout the several views refer to the same or similar components or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments disclosed in the present application and should not be regarded as limiting the scope of the present application.

[0059] Figure 1 is a flowchart of the method for reconstructing the atrial and ventricular defect imaging model;

[0060] Figure 2 is a flowchart of obtaining the two-dimensional medical image data of the heart region;

[0061] Figure 3It is a flow chart for identifying atrial septal defect and ventricular septal defect regions and extracting feature points based on two-dimensional medical image data using image segmentation technology;

[0062] Figure 4 It is a flow chart for further precisely locating the boundaries of atrial septal defect and ventricular septal defect using an edge detection algorithm based on the preliminary identification results;

[0063] Figure 5 It is a flow chart for marking a pixel as a boundary point if the gradient intensity is greater than the set threshold T;

[0064] Figure 6 It is a flow chart for adjusting the threshold T and re - marking if the number of connected components is less than the expected number N;

[0065] Figure 7 It is a flow chart for the steps of identifying atrial septal defect and ventricular septal defect regions and extracting feature points based on two-dimensional medical image data using image segmentation technology;

[0066] Figure 8 It is a flow chart for the steps of preliminarily identifying atrial septal defect and ventricular septal defect regions using a threshold segmentation method based on the pre - processed image data;

[0067] Figure 9 It is a flow chart for the steps of marking a pixel as part of the atrial septal defect and ventricular septal defect regions if the gray value G is greater than the threshold T;

[0068] Figure 10 It is a flow chart for the steps of calculating the difference Δi = Gi - T between the gray value Gi of each pixel i and the threshold T;

[0069] Figure 11 It is a flow chart for the steps of entering the next step if Δi>0; otherwise, returning to the un - marked state. Specific implementation manners

[0070] The present invention will be described in detail below with reference to the accompanying drawings.

[0071] The following specific examples illustrate the embodiments of the present disclosure. Those skilled in the art can easily understand the other advantages and effects of the present disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present disclosure without creative efforts belong to the scope of protection of the present disclosure.

[0072] It should be noted that the following describes various aspects of embodiments within the scope of the appended claims. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is illustrative only. Based on the present disclosure, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement the device and / or practice the method. Additionally, this device and / or this method can be implemented using other structures and / or functionality in addition to one or more of the aspects described herein.

[0073] It should also be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present disclosure schematically. The components shown in the diagrams only include those related to the present disclosure, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in its actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0074] In addition, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0075] Next, refer to Figure 1 Describe the method for reconstructing the atrial septal defect and ventricular septal defect imaging model:

[0076] S101: By obtaining two-dimensional medical image data of the heart region, we first need to ensure that the medical images used are of high quality and contain sufficient information to support subsequent analysis and modeling work. This step typically involves using imaging techniques such as echocardiography, computed tomography (CT), or magnetic resonance imaging (MRI) to capture multiple cross-sectional images of the heart. For example, in a specific case, a doctor may use an MRI device to obtain a series of high-resolution cross-sectional images of the heart, which cover the entire heart region, including the atria, ventricles, and any known atrial and ventricular defects. To ensure the quality of the data, some preprocessing steps may also be required, such as removing noise, enhancing contrast, or correcting image distortion. These preprocessing steps are crucial for improving the accuracy and clarity of the final 3D model.

[0077] S102: By using image segmentation techniques based on the two-dimensional medical image data to identify atrial and ventricular defect regions and extract feature points, this process involves using advanced image processing algorithms to automatically or semi-automatically identify abnormal regions in the heart structure. For example, machine learning algorithms such as convolutional neural networks (CNN) can be used to train a model to identify specific types of atrial and ventricular defects, such as atrial septal defect (ASD) or ventricular septal defect (VSD). Once these regions are identified, the next step is to extract key feature points from these regions, which will be used as the basis for constructing the 3D model. The selection of feature points should be able to accurately reflect the geometry and location of atrial and ventricular defects. For example, after identifying an atrial septal defect, multiple feature points can be placed manually or automatically at the defect edge to ensure that these points can precisely depict the boundary of the defect.

[0078] S103: By using 3D reconstruction algorithms to construct a preliminary 3D model of atrial and ventricular defects based on the feature points, this step is a key link in converting two-dimensional image data into a three-dimensional spatial representation. In this process, various 3D reconstruction algorithms can be used, such as surface reconstruction, volume rendering, or hybrid methods. For example, a surface-based reconstruction method can be used to create a detailed 3D mesh model that can accurately reflect the shape and size of atrial and ventricular defects. To construct this preliminary model, all the previously extracted feature points need to be used as input, and interpolation or fitting techniques are used to generate a continuous surface. In addition, time-series two-dimensional image data can be used to simulate the dynamic changes of the heart to obtain a more realistic and comprehensive 3D model.

[0079] S104: Refine the preliminary 3D model through model optimization techniques to improve its accuracy and clarity, which is an important step in ensuring the quality of the final model. Model optimization can include various techniques and methods, such as smoothing, detail enhancement, topology correction, etc. For example, smoothing algorithms can be used to reduce the irregularities on the model surface while maintaining the integrity of key features; detail enhancement techniques can also be applied to highlight the subtle structures of atrial and ventricular defects, such as the irregular shape of the defect edge or the surrounding tissue structure. In addition, to further improve the accuracy of the model, expert knowledge or clinical data can be introduced for model verification and adjustment. For example, if the preliminary model shows that the size of a certain atrial septal defect does not match the actual measurement results, this problem can be corrected by adjusting the model parameters or repositioning the feature points.

[0080] Through the above steps, we can effectively solve the problem of how to improve the accuracy and clarity of the 3D model in the reconstruction of atrial and ventricular defect imaging models. Each step aims to ensure that as much key information as possible is retained and enhanced during the conversion process from the original 2D image data to the final 3D model, thus providing strong support for clinical diagnosis and treatment.

[0081] Next, refer to Figure 2 Describe the specific steps of obtaining 2D medical image data of the heart region in the atrial and ventricular defect imaging model reconstruction method:

[0082] S201: First, preprocess the original medical image data, which includes gray-scale correction and noise removal. Gray-scale correction is to ensure that the gray-scale values in the image can accurately reflect the true situation of the tissue structure, while noise removal is to reduce the random variations or unwanted information in the image, thereby improving the accuracy of subsequent processing. For example, histogram equalization techniques can be used to adjust the contrast of the image, and Gaussian filters can be applied to smooth the image and reduce the influence of noise.

[0083] S202: Next, based on the preprocessed medical image data, use a specific heart region localization algorithm to identify the heart contour. In this step, techniques such as edge detection and threshold segmentation are usually adopted in the prior art to automatically identify the boundary of the heart. For example, the Canny edge detection algorithm can be used to extract the edge information of the heart contour, and then morphological operations such as dilation and erosion can be combined to further refine the contour to ensure that the heart region is accurately identified.

[0084] The technical solution of the present invention is based on an active contour model (ACM) and an adaptive region growing algorithm for heart region localization to identify the heart contour. The main improvement lies in that the ACM can automatically contract along the regions with larger gradients (such as the heart boundary) to accurately fit the heart contour. The adaptive region growing first grows based on the seed point region to find the approximate contour of the heart, avoiding the ACM falling into local optima and enhancing the localization ability of the heart region. In addition, the ACM may fail to contract in regions with unclear edges, so gradient direction optimization is added to ensure that the curve can also converge correctly in low-contrast regions. The specific process is as follows:

[0085] (1) Adaptive region growing

[0086] Select seed points: Calculate the average gray value I of the heart region avg , and select the pixels close to this value as seed points. The Otsu threshold method is used to ensure that the initial seed points are located inside the heart region.

[0087] Region growing: Set the dynamic growth threshold: T = μ + k·σ, where μ is the mean of the current region, σ is the standard deviation, and k is a dynamically adjustable factor (which can decrease with iteration). According to the threshold T, adjacent pixels are incorporated into the heart region until the growth stagnates.

[0088] (2) Optimize the boundary with the active contour model (ACM)

[0089] Define the ACM energy function:

[0090] E = E internal + E external

[0091] Internal energy E internal : Control the smoothness of the curve and prevent the curve from over-contracting.

[0092] External energy E external : Guide the curve to move along the region with a larger image gradient to match the heart boundary.

[0093] Optimize the ACM contour:

[0094] Use the gradient descent method for optimization to ensure that the curve contracts along the maximum gradient direction (the heart boundary).

[0095] Adopt a non-linear adaptive inertia weight to avoid local optima:

[0096]

[0097] Among them, w min 、w max are respectively the minimum and maximum values of the inertia weight, E min 、E maxare the minimum and maximum values of the energy function respectively, and λ controls the weight decay rate; update the ACM contour C(s) using a non - linear adaptive inertia weight:

[0098]

[0099] Check the convergence condition. If ‖C t+1 - C t ‖ < ∈ (a very small variation), stop the iteration and output the segmented heart region C.

[0100] Handle the problem of edge breakage:

[0101] Calculate the tangential direction of the ACM curve in the region with a lower gradient and perform local interpolation to make the contour continuous.

[0102] S203: Use the heart contour information to crop the image segment containing the heart region. This step can be achieved by determining the minimum bounding rectangle of the heart contour, thus accurately framing the heart region. For example, after identifying the heart contour, calculate its minimum bounding rectangle and use it as the cropping region to obtain an image segment containing only the heart part.

[0103] S204: Finally, perform enhancement processing on the cropped image segment to highlight the atrial septal defect and ventricular septal defect regions. The enhancement processing may include techniques such as contrast enhancement and sharpening to improve the visibility of the atrial septal defect and ventricular septal defect regions. For example, a local contrast enhancement method can be used to increase the contrast between the atrial septal defect and ventricular septal defect regions and the surrounding tissues, or the Laplacian operator can be used for sharpening to make the defect regions more clearly distinguishable.

[0104] Next, refer to Figure 3 Describe the specific steps of using image segmentation technology to identify the atrial septal defect and ventricular septal defect regions and extract feature points in the reconstruction method of the atrial septal defect and ventricular septal defect image model:

[0105] S301: Perform preliminary processing on the two - dimensional medical image data by applying threshold segmentation technology to identify the approximate regions of the atrial septal defect and ventricular septal defect. This process usually involves selecting one or more appropriate thresholds to distinguish the atrial septal defect and ventricular septal defect regions from other tissue structures. For example, in a grayscale image, a grayscale threshold can be selected, and pixels below this threshold are regarded as part of the atrial septal defect and ventricular septal defect regions, thus achieving preliminary region identification.

[0106] S302: On the basis of the preliminary identification, use an edge detection algorithm to further accurately locate the boundaries of the atrial septal defect and ventricular septal defect.

[0107] In the prior art, Canny edge detection is usually used to implement edge detection. However, cardiac images often have low contrast, are blurred, and have a large amount of noise, so that conventional edge detection algorithms cannot obtain good results. This application designs a block potential edge detection algorithm. Based on the preliminary recognition results, the edge detection algorithm is used to further accurately locate the boundaries of atrial defects and ventricular defects, including:

[0108] Assume that the gray value of the target area is approximately uniform, and the curve is shrunk or expanded by the gray mean difference between the inside and outside of the block. The calculation formula of the block potential E(C) is as follows:

[0109]

[0110] where I(x, y) is the gray value of the image at the position (x, y), C is the boundary curve of the current segmentation; c1 and c2 are the average gray values of the inside and outside of the segmentation respectively; μ controls the smoothness of the boundary curve, and the larger the value, the smoother the curve; λ1 and λ2 control the weights of the block potential in the segmentation process.

[0111] Find a curve C through the block potential algorithm, so that the average gray value of the inner block divided by the curve is as close as possible to c1, and the average gray value of the outer block divided by the curve is as close as possible to c2. Since I(x, y) represents the actual pixel gray value, it is still effective in the case of large noise or blurred edges.

[0112] S303: If it is found that there is a large difference between the result of edge detection and the result obtained by threshold segmentation, that is, the overlap rate between the two is less than a predetermined ratio, then the threshold needs to be adjusted and the threshold segmentation step needs to be re-executed. This adjustment can be achieved by increasing or decreasing the threshold until a satisfactory overlap rate is reached. For example, if the overlap rate between the initial threshold segmentation and the subsequent edge detection result is only 60%, and the preset minimum overlap rate is 80%, then the threshold needs to be adjusted until the condition is met.

[0113] S304: After finally determining the atrial defect and ventricular defect regions, key feature points are extracted from within this region. These feature points can be the key position information of the atrial defect and ventricular defect, such as the maximum width of the defect, the center point, etc., for subsequent three-dimensional reconstruction or other analysis and processing. For example, several representative points can be selected from the finally determined atrial defect and ventricular defect regions as feature points, and these points may correspond to the widest part of the defect or specific positions on the defect edge, for the convenience of subsequent analysis and modeling.

[0114] Next, refer to Figure 4 Describe the steps of another method different from the above block potential edge detection algorithm for further accurately locating the boundaries of atrial defects and ventricular defects using the edge detection algorithm according to the preliminary recognition results:

[0115] S401: Smooth the image using a Gaussian filter to reduce the influence of noise. At this stage, it is first necessary to preprocess the original image data to reduce the impact of noise on subsequent analysis. The Gaussian filter is a linear smoothing filter that multiplies and sums the Gaussian kernel with each pixel value in the original image through a convolution operation to obtain the smoothed image. This can effectively remove or reduce random noise in the image while trying to keep the detailed information in the atrial and ventricular defect regions of the image undamaged.

[0116] S402: Calculate the gradient intensity by applying the Sobel operator based on the smoothed image. After the image is smoothed, the Sobel operator is then used to calculate the gradients of each pixel point in the horizontal and vertical directions. The Sobel operator consists of two 3x3 matrices, which are used to detect edges in the horizontal and vertical directions respectively. By applying these two matrices to the smoothed image, the gradient intensity of each pixel point can be obtained, which can then reflect whether that position is likely to belong to the boundary of the atrial and ventricular defects.

[0117] S403: If the gradient intensity is greater than the set threshold T, mark the pixel as a boundary point. After calculating the gradient intensity of each pixel point, a threshold T needs to be set to distinguish which pixel points are more likely to be the boundaries of the atrial and ventricular defects. Generally, the gradient intensity at the boundary is significantly higher than that in the non-boundary region. Therefore, if the gradient intensity of a certain pixel point exceeds the pre-set threshold T, it is considered that the pixel point is located on the boundary of the atrial and ventricular defects and is marked.

[0118] S404: Connect the continuous boundary points to form the complete boundaries of the atrial and ventricular defects. The last step is to connect all the pixels marked as boundary points to form a closed contour, which is the boundary of the atrial and ventricular defects. This can be achieved by tracing adjacent boundary points and connecting them into a path. In this way, the positions and shapes of the atrial and ventricular defects can be accurately located, providing important imaging basis for subsequent diagnosis and treatment.

[0119] Next, refer to Figure 5 Describe the specific steps regarding boundary point detection and connected component analysis in the method for reconstructing the atrial and ventricular defect image model:

[0120] S501: Identify boundary points by calculating the gradient intensity

[0121] First, for each pixel point in the input cardiac image data, an appropriate edge detection algorithm such as the Sobel operator or Canny edge detection is used to calculate the gradient intensity G of that pixel point. This process helps us identify regions in the image with significant intensity changes, which usually correspond to the boundaries between different parts of the cardiac structure.

[0122] S502: Mark boundary points through threshold processing

[0123] Next, a threshold T is set to distinguish which pixel points have a gradient intensity large enough to be marked as boundary points. Specifically, if the gradient intensity G of a certain pixel point is greater than the threshold T, then that pixel point is considered to belong to the boundary and is marked as a boundary point. This step helps extract the boundary information of the cardiac structure from the calculated gradient intensity map.

[0124] S503: Determine the boundary through connected component analysis

[0125] After marking all boundary points, connected component analysis technology is further used to identify and separate different boundary regions. Connected component analysis is an image processing technique that helps us identify connected regions composed of adjacent boundary points. In this way, the boundaries of different structures inside the heart can be determined.

[0126] S504: Optimize the boundary detection result through threshold adjustment

[0127] Finally, check whether the number of connected components obtained through the above steps meets the expected number N. If the condition is not satisfied, that is, the number of connected components is less than the expected number N, then the threshold T needs to be adjusted, and S502 and S503 are repeated until a satisfactory boundary detection result is obtained. This method of dynamically adjusting the threshold can ensure accurate boundary detection even in different cardiac image data.

[0128] Next, refer to Figure 6 Describe the specific steps of adjusting the threshold T and re - marking when the number of connected components in the image model reconstruction method for atrial and ventricular defects is insufficient:

[0129] S601: Evaluate whether the threshold T needs to be adjusted by calculating the number M of current connected components. This process involves analyzing the gradient intensity distribution in the image to identify the connected regions, i.e., connected components, defined by a specific threshold T. If the number M of these connected components is less than the expected number N, it indicates that the current threshold T may be too high, resulting in some actually existing structures not being correctly marked.

[0130] S602: If the calculated number of connected components M is indeed less than the expected number N, then decrease the threshold T and return to recalculate the gradient intensity. This step aims to increase the sensitivity of marking by reducing the threshold T, so that areas that were not originally marked can be identified, increasing the number of connected components. Specifically, the value of the threshold T can be gradually decreased, and after each decrease, the process of recalculating the gradient intensity and marking the connected components is performed again until the number of connected components reaches or exceeds the expected number N.

[0131] S603: If the number of connected components M is greater than or equal to the expected number N after the above steps, then keep the threshold T unchanged. This means that the current threshold T is already sufficient to identify all the expected connected components and no further adjustment is needed. At this time, it can be considered that the threshold T has been optimized to an appropriate level and can accurately reflect the boundary information of atrial septal defect and ventricular septal defect.

[0132] S604: Refine the finally determined boundary points to obtain a more accurate boundary. This step usually involves operations such as smoothing and denoising the edges of the marked connected components to ensure that the boundary is clear and continuous, thereby improving the accuracy of the reconstruction model. For example, morphological operations such as opening or closing operations can be used to remove noise points on the boundary, or curve fitting methods can be used to smooth the boundary contour.

[0133] Next, refer to Figure 7 Describe the specific steps of using image segmentation technology to identify the atrial septal defect and ventricular septal defect regions and extract feature points in the atrial septal defect and ventricular septal defect image model reconstruction method:

[0134] S701: First, preprocess the two-dimensional medical image data, which includes noise reduction and smoothing processing. Noise reduction processing can use methods such as median filtering and Gaussian filtering to remove random noise in the image, while smoothing processing can reduce the details and textures of the image through techniques such as mean filtering, thereby improving the accuracy and efficiency of subsequent processing. For example, in actual operation, a median filter can be used to process the original image to eliminate salt-and-pepper noise, and a Gaussian filter can be applied to further smooth the image and reduce the influence of high-frequency noise.

[0135] S702: Next, based on the preprocessed image data, the atrial septal defect and ventricular septal defect regions are initially identified using threshold segmentation. Threshold segmentation is a common image segmentation technique that divides pixel points into different categories by setting one or more thresholds. In this example, appropriate thresholds can be selected according to the gray-scale differences between the atrial septal defect and ventricular septal defect regions and the surrounding tissues to separate the suspected defect regions from the background. For example, a threshold T can be manually set, and all pixels with gray-scale values lower than T are marked as the atrial septal defect and ventricular septal defect regions, while pixels higher than T are regarded as non-defect regions.

[0136] S703: Then, edge detection algorithms are used to further precisely locate the boundaries of the atrial septal defect and ventricular septal defect. Edge detection algorithms can highlight the edge information in the image, which helps to more accurately define the scope of the atrial septal defect and ventricular septal defect. Commonly used edge detection methods include the Sobel operator, Prewitt operator, and Canny edge detection, etc. For example, the Canny edge detection algorithm can be used to detect the edges in the preprocessed image, and the edge detection effect can be optimized by adjusting algorithm parameters such as high and low thresholds to ensure the continuity and accuracy of the edges.

[0137] S704: Finally, multiple key feature points are selected on the located boundaries for subsequent 3D reconstruction. The selection of these feature points should take into account the geometric shapes and position information of the atrial septal defect and ventricular septal defect so that the actual morphology of the defect can be accurately reflected during the 3D reconstruction process. For example, feature points can be selected at the inflection points of the defect boundary, or feature points can be evenly distributed according to the size and shape of the defect to ensure the accuracy and reliability of the reconstructed model.

[0138] Next, refer to Figure 8 Describe the specific steps for initially identifying the atrial septal defect and ventricular septal defect regions in the method for reconstructing the imaging model of atrial septal defect and ventricular septal defect:

[0139] S801: This process starts by calculating the gray-scale value of each pixel in the image data. The gray-scale value is a standard for measuring the brightness of pixels in an image. For two-dimensional medical images, the gray-scale value of each pixel point reflects the density or light transmittance of the tissue at that position. This step usually involves reading the original image data and applying a specific algorithm to each pixel to determine its gray-scale value.

[0140] S802: These gray-scale values are further processed by determining a threshold T based on the gray-scale histogram. The gray-scale histogram shows the distribution of different gray levels in the image. By analyzing this histogram, a suitable threshold T can be found such that pixels with gray-scale values greater than T are considered to belong to the atrial septal defect and ventricular septal defect regions. The process of selecting the threshold may need to consider multiple factors, such as background noise, tissue contrast, etc., to ensure the accurate separation of the target region.

[0141] S803: If the grayscale value G is greater than the threshold T, then mark the pixel as part of the atrial defect and ventricular defect regions. This step involves traversing all the pixels in the entire image and making marks based on the comparison result between their grayscale values and the threshold T. The marking can be achieved by setting a binary flag. For example, if the grayscale value is greater than the threshold, set the flag to 1 indicating it belongs to the atrial defect and ventricular defect regions, otherwise set it to 0.

[0142] S804: Finally, connect the marked pixels to form the preliminary outlines of the atrial defect and ventricular defect regions. This step typically involves using a connected component analysis algorithm, which can identify and combine adjacent marked pixels to form a continuous boundary, namely the outline of the atrial defect and ventricular defect regions. This connectivity analysis helps to remove isolated noise points and ensure that the identified regions have reasonable geometric shapes and sizes.

[0143] Next, refer to Figure 9 Describe the specific steps for marking the atrial defect and ventricular defect regions in the method for reconstructing the atrial defect and ventricular defect imaging model:

[0144] S901: Determine whether to mark a pixel by calculating the difference between the pixel grayscale value and the threshold - First, for each pixel point i in the image, calculate the difference Δi = Gi - T between its grayscale value Gi and the preset threshold T. This calculation process is used to quantify the positional relationship of the pixel point relative to the threshold, thus providing a basis for subsequent marking.

[0145] S902: Decide whether to mark a pixel by comparing the difference with zero - Then, based on the calculated difference Δi, determine whether it is greater than zero. If Δi > 0, that is, the grayscale value of pixel point i is higher than the threshold T, then mark this pixel point as part of the atrial defect and ventricular defect regions; conversely, if Δi ≤ 0, do not mark it, indicating that this pixel point does not belong to the atrial defect and ventricular defect regions.

[0146] S903: Update the outlines of the atrial defect and ventricular defect regions based on the marking results - Finally, according to the results of the above marking process, update the outline information of the atrial defect and ventricular defect regions. This step ensures that the boundaries of the atrial defect and ventricular defect regions can accurately reflect the defective parts identified in the image and provide precise positioning information for subsequent analysis or processing.

[0147] Through the above steps, the atrial defect and ventricular defect regions in the cardiac image can be effectively identified and marked, providing support for subsequent medical diagnosis.

[0148] Next, refer to Figure 10Describe the specific steps for calculating the difference feature between the gray value of each pixel and the threshold in the method for reconstructing the imaging model of atrial and ventricular defects:

[0149] S1001: First, obtain the gray value Gi of each pixel i in the image. This step can be achieved by reading digital image data, where the gray value of each pixel represents the brightness information at that position. For example, in an 8-bit gray image, the gray value Gi of pixel i ranges from 0 to 255, where 0 represents black and 255 represents white.

[0150] S1002: Then, set a threshold T. The selection of the threshold can be determined according to the specific image content and analysis purpose. For example, when processing cardiac ultrasound images, in order to distinguish atrial and ventricular tissues from surrounding structures, it may be necessary to select a suitable threshold T based on experience or through an automatic algorithm such as the Otsu method.

[0151] S1003: Next, calculate the difference Δi = Gi - T between the gray value Gi of each pixel i and the threshold T. This calculation process can be completed using basic arithmetic operations in a computer programming language. For example, in Python, array operations in the numpy library can be used to efficiently calculate the differences for the entire image.

[0152] S1004: Finally, determine whether the calculated difference Δi is greater than 0. If Δi > 0, it indicates that the gray value of this pixel is higher than the set threshold, meaning that this pixel may be part of the region of interest and can enter the next step for further processing; if Δi ≤ 0, it indicates that the gray value of this pixel is lower than or equal to the threshold, and at this time, the non-marked state is returned, that is, it is considered that this pixel does not belong to the target region and subsequent processing steps can be skipped.

[0153] Through the above steps, the key regions in the images of atrial and ventricular defects can be effectively identified, providing a basis for further model reconstruction.

[0154] Next, refer to Figure 11 Describe the specific steps for pixel marking in the method for reconstructing the imaging model of atrial and ventricular defects:

[0155] S1101: First, calculate the difference value Δi between the current pixel i and its surrounding pixels. This difference value reflects the degree of change in the intensity or texture of pixel i relative to its neighborhood. If the calculated Δi is greater than zero, this indicates that pixel i may belong to the atrial and ventricular defect regions because atrial and ventricular defects usually exhibit different image characteristics from the surrounding normal tissues. At this time, the system will enter the next step, namely the marking process.

[0156] S1102: After determining that Δi is greater than zero, the system marks pixel i as part of the atrial septal defect and ventricular septal defect regions. This marking can be achieved by setting a flag or updating the corresponding position in a two-dimensional array that records whether each pixel in the image is marked as part of the atrial septal defect and ventricular septal defect regions. In this way, it is possible to clearly identify which pixels belong to the atrial septal defect and ventricular septal defect regions.

[0157] S1103: If the calculated Δi is less than or equal to zero, it is considered that pixel i does not belong to the atrial septal defect and ventricular septal defect regions. In this case, the system does not mark pixel i and will continue to check the next pixel to ensure that all pixels in the entire image are evaluated and appropriately marked. This process is implemented through a loop structure to ensure that each pixel is checked and processed one by one.

[0158] During the actual operation process, when this device is used, it is first necessary to obtain two-dimensional medical image data of the patient's heart region, which usually comes from imaging techniques such as echocardiography, CT, or MRI. Subsequently, the system will automatically or assist the doctor in using image segmentation techniques to identify the atrial septal defect and ventricular septal defect regions and accurately extract key feature points from them. This process is crucial for subsequent three-dimensional reconstruction. The extraction of feature points not only depends on the accuracy of the algorithm but also needs to be combined with the doctor's professional knowledge to ensure that the selected feature points can truly reflect the actual situation of the atrial septal defect and ventricular septal defect. Next, using a three-dimensional reconstruction algorithm, a preliminary three-dimensional model of the atrial septal defect and ventricular septal defect is constructed based on the previously extracted feature points. In this step, the algorithm will comprehensively consider the spatial position relationship of all feature points to generate a three-dimensional model that can intuitively display the morphology of the atrial septal defect and ventricular septal defect. Finally, through model optimization techniques, the preliminary constructed three-dimensional model is further refined, including but not limited to smoothing the model surface, enhancing details, and correcting the size of the model, etc., to improve the accuracy and clarity of the three-dimensional model, enabling the doctor to more intuitively and accurately evaluate the specific situation of the atrial septal defect and ventricular septal defect, providing strong support for subsequent diagnosis and treatment. The entire process is a highly integrated process. From data acquisition to the generation of the final model, each link is closely connected, jointly constituting an efficient and accurate method for reconstructing the atrial septal defect and ventricular septal defect imaging model.

[0159] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium.

[0160] In the description of this specification, the descriptions with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0161] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" can explicitly or implicitly include at least one of these features. In the description of the present application, "a plurality of" means two or more unless otherwise specifically defined.

[0162] Any process or method description shown in the flowchart or described in other ways herein can be understood to represent a module, segment, or portion of code including one or more executable instructions for implementing a specific logical function or process. And the scope of the preferred embodiments of the present application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in the reverse order according to the involved functions, rather than in the order shown or discussed.

[0163] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a sequenced list of executable instructions for implementing a logical function, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices.

[0164] It should be understood that each part of the present application can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the method in the above embodiments can be completed by a program instructing relevant hardware. The program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0165] In addition, in each embodiment of the present application, each functional unit can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk, an optical disc, or the like.

[0166] The above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various changes or substitutions, and these should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for reconstructing an imaging model of atrial septal defect and ventricular septal defect, characterized in that, Comprising: Obtain two-dimensional medical image data of the heart region; Based on the two-dimensional medical image data, use image segmentation technology to identify atrial septal defect and ventricular septal defect regions and extract feature points; Use a three-dimensional reconstruction algorithm to construct a preliminary three-dimensional model of atrial septal defect and ventricular septal defect according to the feature points; Refine the preliminary three-dimensional model through model optimization technology to improve its accuracy and clarity; Wherein, the obtaining of the two-dimensional medical image data of the heart region includes: Preprocess the original medical image data, including gray correction and noise removal; Based on the preprocessed medical image data, use a heart region localization algorithm to identify the heart contour; Use the heart contour information to crop out the image segment containing the heart region; Perform enhancement processing on the cropped image segment to highlight the atrial septal defect and ventricular septal defect regions.

2. The method for reconstructing an atrial septal defect and ventricular septal defect imaging model according to claim 1, wherein, Based on the two-dimensional medical image data, using image segmentation technology to identify atrial septal defect and ventricular septal defect regions and extract feature points includes: Apply threshold segmentation technology to preliminarily identify atrial septal defect and ventricular septal defect regions; Based on the preliminary identification result, use an edge detection algorithm to further accurately locate the boundaries of atrial septal defect and ventricular septal defect; If the overlap rate between the edge detection result and the threshold segmentation result is less than a predetermined ratio, adjust the threshold and perform segmentation again; Extract key feature points from the finally determined atrial septal defect and ventricular septal defect regions.

3. The method for reconstructing an atrial septal defect and ventricular septal defect imaging model according to claim 2, wherein Based on the preliminary identification result, using an edge detection algorithm to further accurately locate the boundaries of atrial septal defect and ventricular septal defect includes: The calculation formula for finding the segmentation curve C of the block potential energy E(C) is as follows: Wherein, I(x,y) is the gray value of the image at the position (x,y), C is the boundary curve of the current segmentation; c1 and c2 are the average gray values of the inside and outside of the segmentation respectively; μ controls the smoothness of the boundary curve, the larger the value, the smoother the curve; λ1 and λ2 control the weights of the block potential energy in the segmentation process; find a curve C through the block potential energy algorithm such that the average gray value of the internal block divided by the curve is as close as possible to c1, and the average gray value of the external block divided by the curve is as close as possible to c2.

4. The method for reconstructing an atrial septal defect and ventricular septal defect imaging model according to claim 2, characterized in that, Based on the preliminary identification result, using an edge detection algorithm to further accurately locate the boundaries of atrial septal defect and ventricular septal defect includes: Use a Gaussian filter to smooth the image to reduce the influence of noise; Based on the smoothed image, apply the Sobel operator to calculate the gradient intensity; If the gradient intensity is greater than the set threshold T, mark the pixel as a boundary point; Connect the continuous boundary points to form the complete boundaries of atrial septal defect and ventricular septal defect.

5. The method for reconstructing an atrial septal defect and ventricular septal defect imaging model according to claim 1, characterized in that The steps of based on the two-dimensional medical image data, using image segmentation technology to identify atrial septal defect and ventricular septal defect regions and extract feature points include: Preprocess the two-dimensional medical image data, including noise reduction and smoothing processing; Based on the preprocessed image data, use the threshold segmentation method to preliminarily identify atrial septal defect and ventricular septal defect regions; Use the edge detection algorithm to further accurately locate the boundaries of atrial septal defect and ventricular septal defect; Select multiple key feature points on the located boundary for subsequent three-dimensional reconstruction.

6. The method for reconstructing an atrial septal defect and ventricular septal defect imaging model according to claim 5, wherein The steps of based on the preprocessed image data, using the threshold segmentation method to preliminarily identify atrial septal defect and ventricular septal defect regions include: Calculate the grayscale value of each pixel in the computed image data; Determine a threshold T based on the grayscale histogram such that pixels with grayscale values greater than T belong to the atrial septal defect and ventricular septal defect regions; If the grayscale value G is greater than the threshold T, mark the pixel as part of the atrial septal defect and ventricular septal defect regions; Connect the marked pixels to form a preliminary contour of the atrial septal defect and ventricular septal defect regions.

7. The method for reconstructing atrial septal defect and ventricular septal defect imaging models according to claim 6, wherein The step of if the grayscale value G is greater than the threshold T, mark the pixel as part of the atrial septal defect and ventricular septal defect regions includes: For each pixel i, calculate the difference Δi = Gi - T between its grayscale value Gi and the threshold T; If Δi > 0, mark pixel i as part of the atrial septal defect and ventricular septal defect regions; If Δi ≤ 0, do not mark pixel i; Update the contour of the atrial septal defect and ventricular septal defect regions based on the marking results.

8. The method for reconstructing an atrial septal defect and ventricular septal defect imaging model according to claim 7, characterized in that, The step of for each pixel i, calculate the difference Δi = Gi - T between its grayscale value Gi and the threshold T includes: Obtain the grayscale value Gi of pixel i; Set the threshold T; Calculate the difference Δi = Gi - T; If Δi > 0, proceed to the next step; otherwise return to the unmarked state.

9. The method for reconstructing the atrial septal defect and ventricular septal defect imaging model according to claim 8, characterized in that, The step of if Δi > 0, proceed to the next step; The step of otherwise return to the unmarked state includes: If Δi > 0, mark pixel i as part of the atrial septal defect and ventricular septal defect regions; If Δi ≤ 0, do not mark pixel i and check the next pixel.

Images