A non-contact method and system for three-dimensional cardiac mapping

Abstract

This invention discloses a non-contact cardiac 3D mapping method and system, belonging to the field of image mapping technology. The method involves acquiring a sequence of cardiac ultrasound images across multiple cardiac cycles; preprocessing each cardiac ultrasound image to obtain optimized cardiac ultrasound images, including image enhancement and edge detection; registering each optimized cardiac ultrasound image based on a geometric constraint matching algorithm to ensure they are in the same coordinate system; performing 3D reconstruction using a triangulation algorithm based on the registered optimized cardiac ultrasound images to obtain a 3D cardiac model; performing 3D mapping of the cardiac structure on the 3D cardiac model; and outputting the 3D cardiac mapping results. The use of automated algorithms to process cardiac ultrasound images generates a high-quality 3D cardiac model for mapping in a shorter time, improving operational efficiency and accuracy compared to traditional manual operations, avoiding the subjective influence of manual mapping, and reducing surgical safety risks.

Description

Technical Field

[0001] This invention belongs to the field of image mapping technology, specifically relating to a non-contact three-dimensional cardiac mapping method and system. Background Technology

[0002] Three-dimensional cardiac mapping refers to the method of measuring and analyzing the heart using three-dimensional imaging technology. Through three-dimensional cardiac mapping, doctors can obtain more precise information about the heart's size, shape, volume, and function, thereby enabling more accurate diagnosis and treatment planning for heart conditions. The development of three-dimensional cardiac mapping technology has also provided more information support for cardiac surgery, helping doctors to better plan and perform surgeries.

[0003] Currently, the treatment of relatively complex arrhythmias often requires interventional cardiac catheterization. To improve the success rate of this procedure, preoperative three-dimensional mapping of the patient's heart is essential. However, current three-dimensional mapping of the heart often involves inserting a mapping catheter into the body, which demands a high level of skill from the operator, increases patient discomfort, and, due to varying operator skill levels, often results in highly subjective mapping information, thus posing significant safety risks to the subsequent surgical procedure. Summary of the Invention

[0004] The purpose of this invention is to provide a non-contact cardiac three-dimensional mapping method and system, which can solve the technical problems that existing three-dimensional mapping of the heart often involves the use of mapping catheters to intervene in the human body, which requires a high level of technical skill from the operator, increases the patient's pain, and the mapping results are often highly subjective due to the different skill levels of the operators, thus bringing great safety risks to the subsequent surgical process.

[0005] To solve the above-mentioned technical problems, the present invention is implemented as follows:

[0006] This invention provides a non-contact three-dimensional cardiac mapping method, the method comprising:

[0007] S101: Acquire cardiac ultrasound image sequences within multiple cardiac cycles, wherein the cardiac ultrasound image sequences include multiple cardiac ultrasound images;

[0008] S102: Preprocess each cardiac ultrasound image to obtain an optimized cardiac ultrasound image, wherein the preprocessing includes image enhancement and edge detection;

[0009] S103: Register each optimized cardiac ultrasound image based on the geometric constraint matching algorithm to ensure that each optimized cardiac ultrasound image is in the same coordinate system;

[0010] S104: The triangulation algorithm is used to perform three-dimensional reconstruction based on each registered optimized cardiac ultrasound image to obtain a three-dimensional cardiac model.

[0011] S105: Perform three-dimensional mapping of cardiac structure on a three-dimensional model of the heart;

[0012] S106: Outputs the three-dimensional mapping results of the heart.

[0013] This invention provides a non-contact cardiac three-dimensional mapping system, comprising:

[0014] The acquisition module is used to acquire cardiac ultrasound image sequences over multiple cardiac cycles, wherein the cardiac ultrasound image sequence includes multiple cardiac ultrasound images;

[0015] The preprocessing module is used to preprocess each cardiac ultrasound image to obtain an optimized cardiac ultrasound image, wherein the preprocessing includes image enhancement and edge detection;

[0016] The registration module is used to register each optimized echocardiogram image based on a geometric constraint matching algorithm to ensure that each optimized echocardiogram image is in the same coordinate system.

[0017] The reconstruction module is used to perform three-dimensional reconstruction based on each registered optimized echocardiogram image using a triangulation algorithm to obtain a three-dimensional model of the heart.

[0018] The mapping module is used to perform three-dimensional mapping of the heart structure on a three-dimensional heart model.

[0019] The output module is used to output the three-dimensional mapping results of the heart.

[0020] In this embodiment of the invention, image enhancement is performed on the acquired patient echocardiogram image sequence to increase the clarity of the displayed echocardiogram images. Then, edge detection is performed to obtain the edges of the heart, removing irrelevant factors from the images and avoiding the influence of irrelevant data on the mapping process, thus improving the accuracy of cardiac 3D reconstruction and consequently the accuracy of the cardiac 3D mapping results. Furthermore, the preprocessed optimized echocardiogram images are registered to ensure that each optimized ultrasound image is in the same coordinate system, avoiding inaccuracies in the cardiac 3D model caused by direct reconstruction, further improving the accuracy of the established cardiac 3D model. In addition, this method uses automated algorithms to process the echocardiogram images, generating high-quality cardiac 3D models for mapping in a shorter time. Compared to traditional manual operation, this improves operational efficiency and accuracy, avoids the subjective influence of manual mapping, and reduces surgical safety risks. The entire cardiac 3D mapping process is non-contact, enhancing the patient's treatment experience. Attached Figure Description

[0021] Figure 1This is a flowchart illustrating a non-contact three-dimensional cardiac mapping method provided in an embodiment of the present invention;

[0022] Figure 2 This is a schematic diagram of a non-contact cardiac three-dimensional mapping system provided in an embodiment of the present invention.

[0023] The realization of the objective, functional characteristics and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0025] The following description, in conjunction with the accompanying drawings, details a non-contact cardiac three-dimensional mapping method and system provided by the present invention through specific embodiments and application scenarios. Example 1

[0026] Reference Figure 1 The diagram shows a flowchart of a non-contact cardiac three-dimensional mapping method provided by an embodiment of the present invention.

[0027] This invention provides a non-contact three-dimensional cardiac mapping method, the method comprising:

[0028] S101: Acquire cardiac ultrasound image sequences within multiple cardiac cycles, wherein the cardiac ultrasound image sequences include multiple cardiac ultrasound images.

[0029] Among them, cardiac ultrasound images are a series of two-dimensional images formed by scanning the heart with ultrasound waves. Multiple cardiac ultrasound images can be used to observe and record the heart from different directions and positions.

[0030] It should be noted that because the structure and function of the heart differ in different cardiac cycles, acquiring echocardiogram sequences from multiple cardiac cycles can provide more comprehensive and accurate cardiac information.

[0031] S102: Preprocess each cardiac ultrasound image to obtain an optimized cardiac ultrasound image, wherein the preprocessing includes image enhancement and edge detection.

[0032] It should be noted that image enhancement can make cardiac structures clearer in echocardiogram images, which is beneficial for subsequent image registration and 3D reconstruction. Edge detection can determine the boundary positions of cardiac structures by detecting edge information in the image, avoiding the influence of irrelevant factors on the subsequent 3D reconstruction process.

[0033] In one possible implementation, S1021 specifically includes:

[0034] S1021: Image enhancement is performed on each cardiac ultrasound image through histogram equalization to obtain enhanced images. G ( x , y ):

[0035]

[0036] in, G ( x , y () represents the pixel value of the enhanced image. H ( i , j ) represents the coordinates in the original image ( i , j The pixel value at ) M and N These represent the width and height of the image, respectively.

[0037] Histogram equalization, in particular, adjusts the histogram of pixel grayscale value distribution in echocardiogram images, making the grayscale value distribution more uniform and thus enhancing the visual appeal of the image. In echocardiogram image processing, histogram equalization helps physicians observe cardiac structures and abnormalities more clearly in the images, improving diagnostic accuracy.

[0038] S1022: Perform Sobel edge detection on the enhanced image to extract edge information. G x and G y :

[0039]

[0040]

[0041] in, G x and G y These represent the gradients in the horizontal and vertical directions of the enhanced image after edge detection, respectively. The symbol * indicates a convolution operation. I This indicates an enhanced image.

[0042] It should be noted that the grayscale values ​​of the enhanced cardiac ultrasound images are more uniform, which enhances the visual effect and differences of various elements in the cardiac ultrasound images. The Sobel operator is used for edge detection to detect the grayscale differences between different regions in the image, thereby segmenting the image into different regions, finding the edge positions in the image, and further helping doctors analyze the cardiac structure and abnormalities in the image.

[0043] S1023: Calculate the optimized echocardiogram image based on the gradients in the horizontal and vertical directions of the enhanced image after edge detection. G :

[0044] .

[0045] S1024: Output optimized cardiac ultrasound images G .

[0046] S103: Register each optimized cardiac ultrasound image based on the geometric constraint matching algorithm to ensure that each optimized cardiac ultrasound image is in the same coordinate system.

[0047] It should be noted that the acquired cardiac ultrasound image sequences usually have certain geometric changes and deviations due to differences in space or time, as well as the influence of factors such as the patient's breathing and heart rate. In order to merge these image sequences or use them for subsequent analysis and processing, these images need to be registered to place them in the same coordinate system, so as to ensure that the three-dimensional cardiac images in the subsequent reconstruction process can be seamlessly connected and eliminate splicing marks.

[0048] In one possible implementation, S103 specifically includes:

[0049] S1031: Select one image from multiple optimized echocardiogram images as a reference image, extract feature points from the reference image based on the Fast corner detection method, and obtain a set of feature points for the reference image. Select an image to be aligned from the remaining optimized echocardiogram images, extract feature points from the image to be aligned based on the Fast corner detection method, and obtain a set of feature points for the image to be aligned.

[0050] S1032: Select a reference feature point from the set of feature points in the reference image, and calculate the Euclidean distance between the reference feature point and the feature points in the set of feature points in the image to be aligned. :

[0051]

[0052] in, Represents the coordinates of the reference feature point. This represents the coordinates of feature points in the feature point set of the image to be aligned. i =1,2,..., m , m This indicates the number of feature points in the set of feature points in the image to be aligned.

[0053] S1033: Find the feature point that is closest to the reference feature point in the set of feature points of the image to be aligned, and record the correspondence between the reference feature point and the nearest feature point.

[0054] S1034: Based on the correspondence between the reference feature points and the nearest feature points, the similarity transformation matrix T between the reference image and the image to be aligned is calculated using the least squares method.

[0055]

[0056]

[0057]

[0058] in, Indicates the nearest feature point, s Indicates the scaling factor. θ Indicates the rotation angle, ( t x , t y ) represents the translation vector.

[0059] S1035: Register the images to be aligned so that they are aligned with the reference image.

[0060]

[0061] in,( x , y () represents the pixel coordinates in the image to be aligned. This represents the coordinates of a pixel in the image to be aligned with the reference image, where 1 represents a parameter added for ease of calculation.

[0062] It should be noted that the process of registering images to be aligned mainly involves transforming the images to be aligned into the coordinate system of the reference image so that they can be compared and analyzed in the same coordinate system.

[0063] S104: The triangulation algorithm is used to perform three-dimensional reconstruction based on the registered optimized cardiac ultrasound images to obtain a three-dimensional cardiac model.

[0064] It should be noted that the registered and optimized echocardiogram images are in the same coordinate system. Using the triangulation algorithm, the coordinate positions of these pixels in three-dimensional space are calculated based on the pixel coordinates in the registered and optimized echocardiogram images. A three-dimensional model is then constructed using these points. The points in the model are connected by triangles to form a mesh model composed of many triangles.

[0065] In one possible implementation, S104 specifically includes:

[0066] S1041: Extract feature points from the registered optimized cardiac ultrasound image, including ventricular wall feature points and heart valve feature points, to obtain a feature point set.

[0067] S1042: Convert the set of feature points into a 3D point cloud data set. X :

[0068]

[0069] in, For feature points in the feature point set q i The three-dimensional coordinates.

[0070] S1043: Converting 3D point cloud data into triangular mesh data using a triangulation algorithm. M :

[0071]

[0072] Each triangle m i Corresponding to three 3D point cloud data .

[0073] S1044: Select different materials and textures to optimize and compress the triangular mesh data to obtain a 3D model of the heart.

[0074] S105: Perform three-dimensional mapping of the heart structure on a three-dimensional model of the heart.

[0075] In one possible implementation, S105 specifically includes:

[0076] S1051: Using computer vision algorithms, surface segmentation is performed on the 3D model of the heart to obtain a surface model of the heart structure.

[0077] Alternatively, the computer vision algorithm can use the Sobel operator edge detection algorithm used earlier to segment the established 3D heart model according to the required angle.

[0078] S1052: Preprocess the surface model to detect heart feature points and mark key structures among the heart feature points to obtain a set of structural feature points. The preprocessing includes image enhancement and edge segmentation.

[0079] S1053: Calculate the required three-dimensional cardiac mapping results based on the set of structural feature points.

[0080] Understandably, the reconstructed 3D heart model is built from multiple images that have undergone image enhancement and edge detection processing. Therefore, the surface and internal structures of the reconstructed 3D heart model are displayed very clearly. In actual operation, we may need to obtain the surface model of the heart from different angles. At this time, we can select and segment as needed. The resulting surface model is also clearly displayed and easy to observe. The segmented surface model, combined with the heart's morphological structure and marked structural feature points, is used to calculate the corresponding dimensions or other necessary data, and output them for surgical observation of the patient's internal heart structure, thereby improving the success rate of the surgery.

[0081] S106: Outputs the three-dimensional mapping results of the heart.

[0082] In this embodiment of the invention, image enhancement is performed on the acquired patient echocardiogram image sequence to increase the clarity of the displayed echocardiogram images. Then, edge detection is performed to obtain the edges of the heart, removing irrelevant factors from the images and avoiding the influence of irrelevant data on the mapping process, thus improving the accuracy of cardiac 3D reconstruction and consequently the accuracy of the cardiac 3D mapping results. Furthermore, the preprocessed optimized echocardiogram images are registered to ensure that each optimized ultrasound image is in the same coordinate system, avoiding inaccuracies in the cardiac 3D model caused by direct reconstruction, further improving the accuracy of the established cardiac 3D model. In addition, this method uses automated algorithms to process the echocardiogram images, generating high-quality cardiac 3D models for mapping in a shorter time. Compared to traditional manual operation, this improves operational efficiency and accuracy, avoids the subjective influence of manual mapping, and reduces surgical safety risks. The entire cardiac 3D mapping process is non-contact, enhancing the patient's treatment experience. Example 2

[0083] Reference Figure 2 The diagram shows a structural schematic of a non-contact cardiac three-dimensional mapping system provided in an embodiment of the present invention.

[0084] A non-contact cardiac three-dimensional mapping system 20, comprising:

[0085] Acquisition module 201 is used to acquire cardiac ultrasound image sequences within multiple cardiac cycles, wherein the cardiac ultrasound image sequence includes multiple cardiac ultrasound images.

[0086] Preprocessing module 202 is used to preprocess each of the cardiac ultrasound images to obtain optimized cardiac ultrasound images, wherein the preprocessing includes image enhancement and edge detection;

[0087] Registration module 203 is used to register each of the optimized cardiac ultrasound images based on a geometric constraint matching algorithm to ensure that each of the optimized cardiac ultrasound images is in the same coordinate system;

[0088] The reconstruction module 204 is used to perform three-dimensional reconstruction based on each registered optimized cardiac ultrasound image using a triangulation algorithm to obtain a three-dimensional cardiac model.

[0089] The mapping module 205 is used to perform three-dimensional mapping of the heart structure on the three-dimensional heart model;

[0090] Output module 206 is used to output the three-dimensional mapping results of the heart.

[0091] In one possible implementation, the preprocessing module specifically includes:

[0092] The enhancement submodule is used to enhance each of the cardiac ultrasound images through histogram equalization processing to obtain enhanced images. G ( x , y ):

[0093]

[0094] in, G ( x , y ) represents the pixel value of the enhanced image. H ( i , j ) represents the coordinates in the original image ( i , j The pixel value at ) M and N These represent the width and height of the image, respectively.

[0095] The first extraction submodule is used to perform Sobel edge detection on the enhanced image and extract the edge information of the enhanced image. G x and G y :

[0096]

[0097]

[0098] in, G xand G y These represent the gradients of the enhanced image in the horizontal and vertical directions after edge detection, respectively. The symbol * indicates a convolution operation. I This refers to the enhanced image;

[0099] The first calculation submodule is used to calculate the optimized echocardiogram image based on the gradients in the horizontal and vertical directions of the enhanced image after edge detection. G :

[0100] ;

[0101] The output submodule is used to output the optimized cardiac ultrasound image. G .

[0102] In one possible implementation, the registration module specifically includes:

[0103] The selection submodule is used to select one image as a reference image from multiple optimized cardiac ultrasound images, extract feature points in the reference image based on the Fast corner detection method to obtain a reference image feature point set, select an image to be aligned from the remaining optimized cardiac ultrasound images, extract feature points in the image to be aligned based on the Fast corner detection method to obtain an image feature point set to be aligned.

[0104] The second calculation submodule is used to select a reference feature point from the set of feature points in the reference image and calculate the Euclidean distance between the reference feature point and the feature points in the set of feature points in the image to be aligned. :

[0105]

[0106] in, Represents the coordinates of the reference feature point. This represents the coordinates of the feature points in the set of feature points in the image to be aligned. i =1,2,..., m , m This indicates the number of feature points in the set of feature points in the image to be aligned;

[0107] The recording submodule is used to find the feature point that is closest to the reference feature point in the set of feature points of the image to be aligned, and record the correspondence between the reference feature point and the nearest feature point.

[0108] The third calculation submodule is used to calculate the similarity transformation matrix T between the reference image and the image to be aligned using the least squares method based on the correspondence between the reference feature point and the nearest feature point.

[0109]

[0110]

[0111]

[0112] in, Indicates the nearest feature point, s Indicates the scaling factor. θ Indicates the rotation angle, ( t x , t y ) represents the translation vector;

[0113] The registration submodule registers the image to be aligned so that it is aligned with the reference image.

[0114]

[0115] in,( x , y () represents the pixel coordinates in the image to be aligned. The coordinates of the pixels in the image to be aligned are shown after they are aligned with the reference image, where 1 represents a parameter added for ease of calculation.

[0116] In one possible implementation, the reconstruction module specifically includes:

[0117] The second extraction submodule is used to extract feature points from the registered optimized cardiac ultrasound image, wherein the feature points include ventricular wall feature points and heart valve feature points, to obtain a set of feature points;

[0118] The first conversion submodule is used to convert the feature point set into a three-dimensional point cloud data set. X :

[0119]

[0120] in, For the feature points in the set of feature points q i 3D coordinates;

[0121] The second conversion submodule is used to convert the 3D point cloud data into triangular mesh data using the triangulation algorithm. M :

[0122]

[0123] Each triangle mi Corresponding to three 3D point cloud data ;

[0124] The acquisition submodule is used to select different materials and textures to optimize and compress the triangular mesh data, thereby acquiring the three-dimensional model of the heart.

[0125] In one possible implementation, the calibration module specifically includes:

[0126] The segmentation submodule is used to perform surface segmentation on the three-dimensional model of the heart using computer vision algorithms to obtain a surface model of the heart structure.

[0127] The labeling submodule is used to preprocess the surface model, detect heart feature points, and label key structures among the heart feature points to obtain a set of structural feature points. The preprocessing includes image enhancement and edge segmentation.

[0128] The fourth calculation submodule is used to calculate the required three-dimensional cardiac mapping results based on the set of structural feature points.

[0129] The non-contact cardiac three-dimensional mapping system 20 provided in this embodiment of the invention can realize the various processes implemented in the above method embodiments, and will not be repeated here to avoid repetition.

[0130] In this embodiment of the invention, image enhancement is performed on the acquired patient echocardiogram image sequence to increase the clarity of the displayed echocardiogram images. Then, edge detection is performed to obtain the edges of the heart, removing irrelevant factors from the images and avoiding the influence of irrelevant data on the mapping process, thus improving the accuracy of cardiac 3D reconstruction and consequently the accuracy of the cardiac 3D mapping results. Furthermore, the preprocessed optimized echocardiogram images are registered to ensure that each optimized ultrasound image is in the same coordinate system, avoiding inaccuracies in the cardiac 3D model caused by direct reconstruction, further improving the accuracy of the established cardiac 3D model. In addition, this method uses automated algorithms to process the echocardiogram images, generating high-quality cardiac 3D models for mapping in a shorter time. Compared to traditional manual operation, this improves operational efficiency and accuracy, avoids the subjective influence of manual mapping, and reduces surgical safety risks. The entire cardiac 3D mapping process is non-contact, enhancing the patient's treatment experience.

[0131] The virtual system in this embodiment of the invention can be a system, or a component, integrated circuit, or chip in a terminal.

[0132] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A non-contact three-dimensional mapping method of the heart, characterized in that, The method comprises the following steps: S101: collecting a sequence of cardiac ultrasound images in multiple cardiac cycles, wherein the sequence of cardiac ultrasound images comprises multiple cardiac ultrasound images; S102: preprocessing each of the cardiac ultrasound images to obtain optimized cardiac ultrasound images, wherein the preprocessing comprises image enhancement and edge detection; S103: registering each of the optimized cardiac ultrasound images based on a geometric constraint matching algorithm to ensure that each of the optimized cardiac ultrasound images is in the same coordinate system; S104: performing three-dimensional reconstruction on each of the registered optimized cardiac ultrasound images to obtain a three-dimensional cardiac model by using a triangulation algorithm; S105: performing three-dimensional mapping of a cardiac structure on the three-dimensional cardiac model; S106: outputting the three-dimensional cardiac mapping result; The S105 specifically comprises: S1051: using a computer vision algorithm to perform surface segmentation on the three-dimensional cardiac model to obtain a surface model of the cardiac structure; S1052: preprocessing the surface model to detect cardiac feature points and mark key structures in the cardiac feature points to obtain a set of structural feature points, wherein the preprocessing comprises the image enhancement and the edge segmentation; S1053: calculating the required three-dimensional cardiac mapping result according to the set of structural feature points.

2. The non-contact three-dimensional mapping method of the heart according to claim 1, characterized in that, The S102 specifically comprises: S1021: performing image enhancement on each of the cardiac ultrasound images by histogram equalization to obtain an enhanced image G(x, y): ； wherein G(x, y) represents a pixel value of the enhanced image, H(i, j) represents a pixel value at a coordinate (i, j) in an original image, and M and N represent the width and height of the image, respectively; S1022: performing Sobel operator edge detection on the enhanced image to extract edge information G of the enhanced image x and G y : ； ； wherein G x and G y respectively denote the gradients of the enhanced image in the horizontal and vertical directions after edge detection, the symbol * denotes a convolution operation, and I denotes the enhanced image. S1023: calculating the optimized cardiac ultrasound image G according to the gradients in the horizontal direction and the vertical direction after edge detection on the enhanced image; ； S1024: outputting the optimized cardiac ultrasound image G.

3. The non-contact three-dimensional mapping method of the heart according to claim 1, characterized in that, The S103 specifically comprises: S1031: selecting one of the optimized cardiac ultrasound images as a reference image, extracting feature points in the reference image based on a Fast corner detection method to obtain a set of reference image feature points, selecting a to-be-aligned image from the remaining optimized cardiac ultrasound images, and extracting feature points in the to-be-aligned image based on the Fast corner detection method to obtain a set of to-be-aligned image feature points; S1032: selecting a reference feature point from the reference image feature point set, and calculating the Euclidean distance between the reference feature point and the feature points in the to-be-aligned image feature point set : ； wherein, denotes the reference feature point, denotes a feature point in the set of image feature points to be aligned, i = 1, 2, …, m, m representing the number of feature points in the set of image feature points to be aligned; S1033: finding a feature point in the set of to-be-aligned image feature points that is closest to the reference feature point as a nearest feature point, and recording the correspondence between the reference feature point and the nearest feature point; S1034: calculating a similarity transformation matrix T between the reference image and the to-be-aligned image by using a least squares method according to the correspondence between the reference feature point and the nearest feature point: ； ； ； wherein, denotes the most recent feature point, s denotes a scaling factor, 0 denotes a rotation angle, (t x ,t y ) denotes a translation vector; S1035: registering the to-be-aligned image to align the to-be-aligned image with the reference image: ； wherein (x, y) represents the pixel point coordinate in the image to be aligned, represents the coordinate of the pixel point in the image to be aligned after alignment with the reference image, wherein 1 represents a parameter added for convenience of calculation.

4. The non-contact three-dimensional mapping method of a heart according to claim 1, characterized by, The S104 specifically comprises: S1041: extracting feature points in the registered optimized cardiac ultrasound image, wherein the feature points comprise ventricular wall feature points and cardiac valve feature points to obtain a set of feature points; S1042: convert the feature point set into a three-dimensional point cloud data set X: ； wherein, a three-dimensional coordinate of a feature point in the set of feature points ; S1043: convert the three-dimensional point cloud data into triangular mesh data M using the triangulation algorithm: ； ； ； wherein each triangle m i corresponding to three three-dimensional point cloud data ; S1044: select different materials and maps to optimize and compress the triangular mesh data, and obtain the three-dimensional model of the heart.

5. A non-contact three-dimensional cardiac mapping system, characterized by, Comprise: The acquisition module is used for acquiring a sequence of heart ultrasound images in a plurality of cardiac cycles, wherein the sequence of heart ultrasound images comprises a plurality of heart ultrasound images; The preprocessing module is used for preprocessing each of the heart ultrasound images to obtain an optimized heart ultrasound image, wherein the preprocessing comprises image enhancement and edge detection; The registration module is used for registering each of the optimized heart ultrasound images based on a geometric constraint matching algorithm to ensure that each of the optimized heart ultrasound images is in the same coordinate system; The reconstruction module is used for performing three-dimensional reconstruction on each of the registered optimized heart ultrasound images based on a triangulation algorithm to obtain a three-dimensional model of the heart; The mapping module is used for performing three-dimensional mapping of the heart structure on the three-dimensional model of the heart; The output module is used for outputting the three-dimensional mapping result of the heart; The mapping module specifically comprises: The segmentation sub-module is used for segmenting the surface of the three-dimensional model of the heart using a computer vision algorithm to obtain a surface model of the heart structure; The labeling sub-module is used for preprocessing the surface model, detecting heart feature points, and labeling key structures in the heart feature points to obtain a set of structure feature points, wherein the preprocessing comprises the image enhancement and the edge segmentation; The fourth calculation sub-module is used for calculating the required three-dimensional mapping result of the heart based on the set of structure feature points.

6. The non-contact three-dimensional cardiac mapping system of claim 5, wherein, The preprocessing module specifically comprises: The enhancement sub-module is used for performing image enhancement on each of the heart ultrasound images by histogram equalization to obtain an enhanced image G(x, y): ； Wherein, G(x, y) represents the pixel value of the enhanced image, H(i, j) represents the pixel value of the original image at coordinate (i, j), and M and N represent the width and height of the image, respectively; A first extraction submodule is configured to perform Sobel operator edge detection on the enhanced image to extract edge information G of the enhanced image. x and G y : ； ； wherein G x and G y respectively represent the gradients of the enhanced image in the horizontal and vertical directions after edge detection, the symbol * represents a convolution operation, and I represents the enhanced image. The first calculation sub-module is used for calculating the optimized heart ultrasound image G based on the gradients in the horizontal direction and the vertical direction after edge detection on the enhanced image; ； The output sub-module is used for outputting the optimized heart ultrasound image G.

7. The non-contact three-dimensional cardiac mapping system of claim 6, wherein, The registration module specifically comprises: The selection sub-module is used for selecting one image from the plurality of optimized heart ultrasound images as a reference image, extracting feature points in the reference image based on a Fast corner detection method to obtain a set of reference image feature points, selecting a to-be-aligned image from the remaining optimized heart ultrasound images, and extracting feature points in the to-be-aligned image based on the Fast corner detection method to obtain a set of to-be-aligned image feature points; a second calculating sub-module, configured to select a reference feature point from the reference feature point set, and calculate the Euclidean distance between the reference feature point and a feature point in the feature point set of the image to be aligned : ； wherein, denotes the reference feature point, denotes a feature point in the set of image feature points to be aligned, i = 1, 2, …, m, m representing the number of feature points in the set of image feature points to be aligned; The recording sub-module is used for finding the feature point closest to the reference feature point in the set of to-be-aligned image feature points as the closest feature point, and recording the correspondence between the reference feature point and the closest feature point. a third calculation submodule configured to calculate a similarity transformation matrix T between the reference image and the image to be aligned using a least square method according to the correspondence between the reference feature points and the nearest feature points; ； ； ； wherein, denotes the most recent feature point, s denotes a scaling factor, 0 denotes a rotation angle, (t x ,t y ) denotes a translation vector; a registration submodule configured to register the image to be aligned so as to align the image to be aligned with the reference image; ； wherein (x, y) represents the pixel point coordinate in the image to be aligned, represents the coordinate of the pixel point in the image to be aligned after alignment with the reference image, wherein 1 represents a parameter added for convenience of calculation.

8. The non-contact three-dimensional cardiac mapping system of claim 7, wherein, the reconstruction module specifically comprises: a second extraction submodule configured to extract feature points in the registered and optimized cardiac ultrasound image, wherein the feature points include ventricular wall feature points and cardiac valve feature points, and obtain a feature point set; a first conversion submodule configured to convert the feature point set into a three-dimensional point cloud data set X; ； wherein, a three-dimensional coordinate of a feature point in the set of feature points ; a second conversion submodule configured to convert the three-dimensional point cloud data into triangular mesh data M using the triangulation algorithm; ； ； ； wherein each triangle m i corresponding to three three-dimensional point cloud data ; an acquisition submodule configured to select different materials and maps to optimize and compress the triangular mesh data, and acquire the cardiac three-dimensional model.

Images