A blood vessel naming method and device, electronic equipment and storage medium

Abstract

This application provides a method, apparatus, electronic device, and storage medium for naming blood vessels. The naming method includes: inputting an angiography image into a pre-trained blood vessel classification model to obtain a blood vessel category; inputting the angiography image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image, and extracting point cloud features to obtain point cloud features of foreground points in the blood vessel segmentation image; inputting the point cloud features of the foreground points in the blood vessel segmentation image into a pre-trained point cloud segmentation model corresponding to the blood vessel category to obtain a fine classification result for each foreground point in the blood vessel segmentation image; and establishing the topological structure of the blood vessel based on the fine classification results to obtain the blood vessel naming result. The technical solution provided in this application can avoid noise interference and improve the accuracy of blood vessel naming.

Description

Technical Field

[0001] This application relates to the field of medical imaging technology, and in particular to a method, apparatus, electronic device, and storage medium for naming blood vessels. Background Technology

[0002] Vascular diseases pose a serious threat to human life and health, and invasive angiography remains the gold standard for vascular diseases such as coronary heart disease and stroke. In the intelligent screening process for vascular diseases, accurately naming individual vessels from images is an essential yet challenging task. Accurate vessel naming is of great significance for subsequent plaque and stenosis analysis and is also a crucial part of vessel registration and three-dimensional reconstruction. However, due to the influence of noise, artifacts, and the patient's breathing and heartbeat in the angiographic images, vessel naming remains a significant challenge.

[0003] Currently, most vascular naming techniques use a single model to name blood vessels. However, due to the large amount of occlusion and overlap between blood vessels in angiography images, and the significant differences between angiography images from different angles, the naming of blood vessels using a single model is not accurate enough. Therefore, how to accurately name blood vessels has become an urgent problem to be solved. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a method, device, electronic device and storage medium for naming blood vessels, which can classify blood vessels in angiographic images, train corresponding point cloud segmentation models for different categories of angiographic images, and input images into the point cloud segmentation model with only foreground information to avoid noise interference, thereby improving the accuracy of the fine classification results of different categories of blood vessels. Furthermore, the topological structure of the image is preserved through point cloud features, and the final blood vessel naming result is obtained by combining the topological structure, thus improving the accuracy of blood vessel naming.

[0005] This application mainly includes the following aspects:

[0006] In a first aspect, embodiments of this application provide a method for naming blood vessels, the naming method comprising:

[0007] Obtain angiographic images of blood vessels;

[0008] The angiography image is input into a pre-trained blood vessel classification model to obtain the blood vessel category corresponding to the angiography image;

[0009] The angiography image is input into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiography image. Point cloud features are extracted from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image.

[0010] Based on the blood vessel category corresponding to the angiography image, the point cloud features of the foreground points in the blood vessel segmentation image are input into the pre-trained point cloud segmentation model corresponding to the blood vessel category to obtain the fine classification result of each foreground point in the blood vessel segmentation image;

[0011] Based on the fine classification results of each foreground point in the segmented blood vessel image, the topological structure of the blood vessel is established, and based on the topological structure of the blood vessel, the naming result of the blood vessel is obtained.

[0012] Furthermore, point cloud features are extracted from the blood vessel segmentation image through the following steps to obtain the point cloud features of the foreground points in the blood vessel segmentation image:

[0013] Obtain the predicted label value, location coordinates, and grayscale value of each point cloud in the blood vessel segmentation image;

[0014] Based on the predicted label value of each point cloud, foreground points are extracted from the point cloud of the blood vessel segmentation image.

[0015] The position coordinates of the foreground point in the blood vessel segmentation image are normalized to obtain the normalized coordinates of the foreground point.

[0016] For each foreground point in the blood vessel segmentation image, the position coordinates of the foreground point, the normalized coordinates of the foreground point, and the gray value of the foreground point are concatenated to obtain the point cloud feature of the foreground point in the blood vessel segmentation image.

[0017] Furthermore, the point cloud segmentation model is trained through the following steps:

[0018] Obtain the label of each foreground point in the blood vessel segmentation image;

[0019] For each foreground point in a sample, the point cloud features of the foreground point in the sample are concatenated with the label of the foreground point in the sample to obtain the target features of the foreground point.

[0020] For each foreground point, the target features of that foreground point are input into the point cloud segmentation model corresponding to the blood vessel category to obtain the predicted fine classification result of that foreground point;

[0021] For each foreground point, the predicted sub-classification result of the foreground point and the label of the foreground point are used to calculate the loss value. If the loss value is not less than a preset threshold, the model parameters of the point cloud segmentation model are adjusted to continue training the point cloud segmentation model. If the loss value is less than the preset threshold, the training is completed, and the trained point cloud segmentation model corresponding to the blood vessel category is obtained.

[0022] Furthermore, the step of establishing the topological structure of the blood vessels based on the fine classification results of each foreground point in the blood vessel segmentation image, and obtaining the naming results of the blood vessels based on the topological structure of the blood vessels, includes:

[0023] Based on the fine classification results of each foreground point in the blood vessel segmentation image, the foreground point indicated by the fine classification results as a fine branch category is obtained, and the name corresponding to the fine branch category is determined as the name of the foreground point of the fine branch category;

[0024] The leading point of the main coronary artery category is obtained from the detailed classification results, and the name corresponding to the main coronary artery category is determined as the name of the leading point of the main coronary artery category;

[0025] The foreground points indicating the main branch categories are obtained from the fine classification results. Based on the foreground points of the main branch categories, a topology structure composed of the foreground points of the main branch categories is established, and the name of each main branch in the topology structure is obtained.

[0026] The naming of the precursor points of the minor branch categories, the naming of the precursor points of the major coronary artery categories, and the naming of each major branch are determined as the naming results of the blood vessels.

[0027] Furthermore, the step of establishing a topology structure composed of the foreground points of the main branch categories based on the foreground points of the main branch categories, and obtaining the naming of each main branch in the topology structure, includes:

[0028] Based on the anterior points of the major branch categories, the anterior points of the major branch categories and the anterior points of the major coronary artery categories are determined as target anterior points;

[0029] The target foreground point is subjected to skeletonization processing to obtain the centerline results of the main branches of the blood vessel;

[0030] Determine the number of adjacent pixels for each pixel in the centerline result, and determine the pixel type for each pixel based on the number of adjacent pixels.

[0031] In the point type, the pixels belonging to the endpoints and bifurcation points are determined as target pixels;

[0032] The target pixels are subjected to depth-first processing to obtain a tree-like topology structure composed of the target pixels;

[0033] The tree-like topology is mapped using the anatomical structure of the coronary arteries to obtain the names of each major branch in the tree-like topology.

[0034] Furthermore, the step of determining the point type of each pixel based on the number of adjacent pixels of each pixel includes:

[0035] Based on the number of adjacent pixels of each pixel, if the number of adjacent pixels of each pixel is one, then the pixel type of the pixel is determined to be an endpoint.

[0036] If a pixel is adjacent to two other pixels, then the pixel type is determined to be the middle point.

[0037] If a pixel has more than two adjacent pixels, then the pixel type is determined to be a bifurcation point.

[0038] Secondly, embodiments of this application also provide a blood vessel naming device, the naming device comprising:

[0039] The acquisition module is used to acquire angiographic images of blood vessels;

[0040] The first processing module is used to input the angiography image into a pre-trained blood vessel classification model to obtain the blood vessel category corresponding to the angiography image;

[0041] The second processing module is used to input the angiography image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiography image, and to extract point cloud features from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image.

[0042] The third processing module is used to input the point cloud features of the foreground points in the blood vessel segmentation image into the pre-trained point cloud segmentation model corresponding to the blood vessel category based on the blood vessel category of the angiography image, so as to obtain the fine classification result of each foreground point in the blood vessel segmentation image.

[0043] The naming module is used to establish the topological structure of the blood vessels based on the fine classification results of each foreground point in the segmented blood vessel image, and to obtain the naming result of the blood vessels based on the topological structure of the blood vessels.

[0044] Furthermore, when the second processing module extracts point cloud features from the segmented blood vessel image to obtain the point cloud features of the foreground points in the segmented blood vessel image, the second processing module is specifically used for:

[0045] Obtain the predicted label value, location coordinates, and grayscale value of each point cloud in the blood vessel segmentation image;

[0046] Based on the predicted label value of each point cloud, foreground points are extracted from the point cloud of the blood vessel segmentation image.

[0047] The position coordinates of the foreground point in the blood vessel segmentation image are normalized to obtain the normalized coordinates of the foreground point.

[0048] For each foreground point in the blood vessel segmentation image, the position coordinates of the foreground point, the normalized coordinates of the foreground point, and the gray value of the foreground point are concatenated to obtain the point cloud feature of the foreground point in the blood vessel segmentation image.

[0049] Thirdly, embodiments of this application also provide an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, they perform the steps of the blood vessel naming method described above.

[0050] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the blood vessel naming method described above.

[0051] This application provides a method, apparatus, electronic device, and storage medium for naming blood vessels. The naming method includes: acquiring an angiographic image of a blood vessel; inputting the angiographic image into a pre-trained blood vessel classification model to obtain a blood vessel category corresponding to the angiographic image; inputting the angiographic image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiographic image, and extracting point cloud features from the blood vessel segmentation image to obtain point cloud features of foreground points in the blood vessel segmentation image; based on the blood vessel category corresponding to the angiographic image, inputting the point cloud features of foreground points in the blood vessel segmentation image into a pre-trained point cloud segmentation model corresponding to the blood vessel category to obtain a fine classification result for each foreground point in the blood vessel segmentation image; establishing the topological structure of the blood vessel based on the fine classification result of each foreground point in the blood vessel segmentation image, and obtaining the naming result of the blood vessel based on the topological structure of the blood vessel.

[0052] Thus, the technical solution provided in this application can classify blood vessels in angiographic images, train corresponding point cloud segmentation models for different categories of angiographic images, and input images into the point cloud segmentation model that only have foreground information, which can avoid noise interference, thereby improving the accuracy of the fine classification results for different categories of blood vessels. Furthermore, the topological structure of the image is preserved through point cloud features, and finally, the topological structure is combined to obtain the final blood vessel naming result, which improves the accuracy of blood vessel naming.

[0053] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0054] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 A flowchart illustrating a method for naming blood vessels provided in an embodiment of this application is shown;

[0056] Figure 2 A flowchart illustrating another method for naming blood vessels provided in an embodiment of this application is shown;

[0057] Figure 3 A schematic diagram showing the centerline result of a major branch provided in an embodiment of this application is illustrated;

[0058] Figure 4 A schematic diagram of a topology provided in an embodiment of this application is shown;

[0059] Figure 5 A schematic diagram illustrating a main branch naming scheme provided in an embodiment of this application is shown;

[0060] Figure 6 This illustration shows one of the structural diagrams of a blood vessel naming device provided in an embodiment of this application;

[0061] Figure 7 This is a second structural diagram of a blood vessel naming device provided in an embodiment of this application;

[0062] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0064] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0065] In order to enable those skilled in the art to use the content of this application, and in combination with the specific application scenario of "name of blood vessels", the following implementation is given. For those skilled in the art, the general principles defined herein can be applied to other embodiments and application scenarios without departing from the spirit and scope of this application.

[0066] The methods, apparatus, electronic devices, or computer-readable storage media described in this application can be applied to any scenario where blood vessels need to be named. This application does not limit the specific application scenario. Any scheme that uses a blood vessel naming method, apparatus, electronic device, and storage medium provided in this application is within the protection scope of this application.

[0067] It is worth noting that vascular diseases pose a serious threat to human life and health. Invasive angiography remains the gold standard for vascular diseases such as coronary heart disease and stroke. In the intelligent screening process for vascular diseases, accurately naming individual vessels from images is an essential yet challenging task. Accurate vessel naming is of great significance for subsequent plaque and stenosis analysis and is also a crucial part of vessel registration and three-dimensional reconstruction. However, due to the influence of noise, artifacts, and factors such as patient breathing and heartbeat in the angiographic images, vessel naming remains a significant challenge.

[0068] Currently, most vessel naming techniques use a single model to name vessels. However, due to the significant occlusion and overlap between vessels in angiographic images, and the substantial differences between angiographic images from different angles, the naming using a single model is not accurate enough. In addition, some vessel naming techniques use the entire image as input, introducing background noise interference and ignoring the topological structure between vessels. Therefore, how to accurately name vessels has become an urgent problem to be solved.

[0069] Based on this, this application proposes a method, apparatus, electronic device, and storage medium for naming blood vessels. The naming method includes: acquiring an angiographic image of a blood vessel; inputting the angiographic image into a pre-trained blood vessel classification model to obtain the blood vessel category corresponding to the angiographic image; inputting the angiographic image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiographic image, and extracting point cloud features from the blood vessel segmentation image to obtain point cloud features of foreground points in the blood vessel segmentation image; based on the blood vessel category corresponding to the angiographic image, inputting the point cloud features of foreground points in the blood vessel segmentation image into a pre-trained point cloud segmentation model corresponding to the blood vessel category to obtain a fine classification result for each foreground point in the blood vessel segmentation image; based on the fine classification result for each foreground point in the blood vessel segmentation image, establishing the topological structure of the blood vessel, and obtaining the naming result of the blood vessel based on the topological structure of the blood vessel.

[0070] Thus, the technical solution provided in this application can classify blood vessels in angiographic images, train corresponding point cloud segmentation models for different categories of angiographic images, and input images into the point cloud segmentation model that only have foreground information, which can avoid noise interference, thereby improving the accuracy of the fine classification results for different categories of blood vessels. Furthermore, the topological structure of the image is preserved through point cloud features, and finally, the topological structure is combined to obtain the final blood vessel naming result, which improves the accuracy of blood vessel naming.

[0071] To facilitate understanding of this application, the technical solutions provided in this application will be described in detail below with reference to specific embodiments.

[0072] Please see Figure 1 , Figure 1 A flowchart illustrating a method for naming blood vessels provided in an embodiment of this application is shown below. Figure 1 As shown, the naming method includes:

[0073] S101. Obtain angiographic images of the blood vessels;

[0074] In this step, the angiographic image is the first unprocessed initial image of the blood vessels captured.

[0075] S102. Input the angiography image into a pre-trained blood vessel classification model to obtain the blood vessel category corresponding to the angiography image;

[0076] In this step, the vessel classification model can adopt a classic convolutional neural network (CNN) structure (such as the VGG model or ResNet). Taking coronary angiography images as an example, the vessel categories can be divided into the left anterior descending artery (LAD), the left circumflex artery (LCX), and the right coronary artery (RCA). The category label for LAD can be 0, the category label for LCX can be 1, and the category label for RCA can be 2. Through supervised learning of the vessel classification model, the loss is gradually reduced (such as cross-entropy loss) until the vessel classification model converges, resulting in a trained vessel classification model. When testing the vessel classification model, an angiography image can be input, and the corresponding category label for that angiography image can be output, for example, 1.

[0077] S103. Input the angiography image into a pre-trained blood vessel segmentation model to obtain the blood vessel segmentation image corresponding to the angiography image, and extract point cloud features from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image.

[0078] In this step, the vessel segmentation model can adopt a classic CNN structure (such as the semantic segmentation network U-Net); taking a coronary angiography image as an example, the corresponding binary segmented image of the vessel is used as the label. Through supervised learning of the vessel segmentation model, the loss is continuously reduced iteratively (such as cross-entropy loss, semantic segmentation loss DICE) until the model converges and a trained vessel segmentation model is obtained; when testing the vessel segmentation model, an angiography image can be input, and the output is the binary segmented image of the corresponding vessel (i.e., the vessel segmentation image).

[0079] It should be noted that you should refer to [link / reference]. Figure 2 , Figure 2 A flowchart illustrating another method for naming blood vessels provided in this application embodiment is shown below. Figure 2 As shown, point cloud features are extracted from the blood vessel segmentation image through the following steps to obtain the point cloud features of the foreground points in the blood vessel segmentation image:

[0080] S201. Obtain the predicted label value, location coordinates, and grayscale value of each point cloud in the blood vessel segmentation image;

[0081] S202. Based on the predicted label value of each point cloud, extract the foreground points from the point cloud of the blood vessel segmentation image;

[0082] S203. Normalize the position coordinates of the foreground point in the blood vessel segmentation image to obtain the normalized coordinates of the foreground point;

[0083] S204. For each foreground point in the blood vessel segmentation image, the position coordinates of the foreground point, the normalized coordinates of the foreground point, and the gray value of the foreground point are concatenated to obtain the point cloud feature of the foreground point in the blood vessel segmentation image.

[0084] Steps S201-S204 involve the extraction of point cloud features. Since the blood vessel segmentation image is a binary image, foreground points can be extracted based on the predicted label value of each point cloud in the blood vessel segmentation image. For example, 0 represents background points and 1 represents foreground points. Therefore, the points with a predicted label value of 1 can be extracted from all point clouds in the blood vessel segmentation image; these are the foreground points. Here, the position coordinates [x...] of the foreground points in the blood vessel segmentation image can be extracted. i ,y i (i = 1, ..., n, where n is the number of foreground points in the blood vessel segmentation image), and normalized coordinates are obtained through normalization. (in (The mean coordinates of all selected foreground points); secondly, obtain the grayscale value of the corresponding foreground point in the angiographic image [g]. i Finally, the three features mentioned above are stitched together to form a point cloud feature of the foreground point.

[0085] S104. Based on the blood vessel category corresponding to the angiography image, the point cloud features of the foreground points in the blood vessel segmentation image are input into the pre-trained point cloud segmentation model corresponding to the blood vessel category to obtain the fine classification result of each foreground point in the blood vessel segmentation image.

[0086] In this step, the point cloud segmentation model can employ common deep learning models such as PointNet or PointTransformer (PT). Taking PointTransformer as an example, the point cloud segmentation model can consist of several downsampling (transition down) and upsampling (transition up) modules, with each layer potentially incorporating a self-attention module (Transformer Block). The self-attention module uses vector attention, as shown in the following mathematical formula:

[0087]

[0088] Among them, y i X is the output feature of the i-th foreground point; i X is the input feature of the i-th foreground point; jχ(i) is the input feature of the j-th foreground point; χ(i) is the set of point clouds in the neighborhood of the i-th foreground point; γ is the mapping function, such as MLP; ρ is the normalization exponential function, such as softmax, used to normalize the features; ψ and α are linear mapping layers; δ is the positional encoding function.

[0089] It should be noted that the point cloud segmentation model is trained through the following steps:

[0090] 1. Obtain the label of each foreground point in the sample blood vessel segmentation image;

[0091] 2. For each foreground point in a sample, the point cloud features of the foreground point in the sample are concatenated with the label of the foreground point in the sample to obtain the target features of the foreground point.

[0092] 3. For each foreground point, input the target features of the foreground point into the point cloud segmentation model corresponding to the blood vessel category to obtain the predicted detailed classification result of the foreground point;

[0093] Fourth, for each foreground point, calculate the loss by combining the predicted sub-classification result of the foreground point with the label of the foreground point. If the loss value is not less than a preset threshold, adjust the model parameters of the point cloud segmentation model to continue training the point cloud segmentation model. If the loss value is less than the preset threshold, the training is complete, and a trained point cloud segmentation model corresponding to the blood vessel category is obtained.

[0094] For example, in steps one through four above, the training process of the point cloud segmentation model first involves a professional doctor annotating the sample blood vessel segmentation images. For instance, the annotation rules could be: labeling small branches in the sample blood vessel segmentation images as 0, labeling major coronary arteries (left anterior descending artery LAD / left circumflex artery LCX / right coronary artery RCA) as 1, and labeling major branches (diagonal branch D / obtuse marginal branch OM / right posterior descending artery RPDA) as 2. The annotation corresponding to the i-th foreground point is denoted as l. i This feature is then concatenated with the point cloud features of the input i-th foreground point to obtain the target features of the i-th foreground point. Will As input to the point cloud segmentation model, the point cloud segmentation model learns the features of the point cloud and outputs a prediction result. Compare the output prediction results with the labeled gold standard. iCalculate the loss (the loss function can be cross-entropy loss), and iterate to reduce the loss until the point cloud segmentation model converges, resulting in a trained point cloud segmentation model. Furthermore, Point Transformers can be trained separately for LCA, LCX, and RCA point cloud segmentation models based on the classified angiography images. The testing process for the point cloud segmentation model is as follows: Input the point cloud features of each foreground point in the test vessel segmentation image, and output the corresponding fine classification result (0, 1, 2) for the foreground point. For example, a fine classification result of 1 indicates that the foreground point belongs to a major coronary artery.

[0095] S105. Based on the fine classification results of each foreground point in the blood vessel segmentation image, establish the topological structure of the blood vessel, and based on the topological structure of the blood vessel, obtain the naming result of the blood vessel.

[0096] In this step, by using the fine classification results of each foreground point in the blood vessel segmentation image, the final naming of the foreground points of small branches or major coronary arteries can be obtained. Since there may be multiple major branches, the foreground points of the major branches obtained by the fine classification results still need to be further determined by establishing the topological structure of the blood vessels to determine the final naming of the foreground points of the major branches.

[0097] It should be noted that the steps of establishing the topological structure of blood vessels based on the fine classification results of each foreground point in the blood vessel segmentation image, and obtaining the naming results of blood vessels based on the topological structure, include:

[0098] S1051. Based on the fine classification results of each foreground point in the blood vessel segmentation image, obtain the foreground point indicated by the fine classification results as a fine branch category, and determine the name corresponding to the fine branch category as the name of the foreground point of the fine branch category;

[0099] S1052. Obtain the front point of the main coronary artery category indicated by the fine classification result, and determine the name corresponding to the main coronary artery category as the name of the front point of the main coronary artery category;

[0100] In steps S1051-S1052, taking LAD as an example, after obtaining the fine classification results, the precursor points belonging to minor branches (i.e., small branches) and the precursor points of the main coronary artery LAD can be obtained from the fine classification results. The names of the precursor points belonging to small branches are then assigned to the small branch category, and the names of the precursor points belonging to the main coronary artery are assigned to the main coronary artery category. Next, the naming of the main branches needs to be determined by establishing the topology. Here, the names corresponding to the small branch category and the main coronary artery category can be predefined names with corresponding meanings.

[0101] S1053. Obtain the foreground points indicating the main branch categories from the fine classification results. Based on the foreground points of the main branch categories, establish a topology structure composed of the foreground points of the main branch categories, and obtain the name of each main branch in the topology structure.

[0102] It should be noted that the steps for establishing a topology composed of foreground points of major branch categories, and obtaining the naming of each major branch in the topology, include:

[0103] 1) Based on the anterior points of the major branch categories, the anterior points of the major branch categories and the anterior points of the major coronary artery categories are determined as target anterior points;

[0104] 2) Perform skeletonization processing on the target foreground point to obtain the centerline results of the main branches of the blood vessel;

[0105] 3) Determine the number of adjacent pixels for each pixel in the centerline result, and determine the pixel type for each pixel based on the number of adjacent pixels for each pixel;

[0106] It should be noted that the steps for determining the point type of each pixel based on the number of its neighboring pixels include:

[0107] (1) Based on the number of adjacent pixels of each pixel, for each pixel, if the number of adjacent pixels of the pixel is one, the point type of the pixel is determined to be an endpoint;

[0108] (2) If the number of adjacent pixels of a pixel is two, then the pixel type of the pixel is determined to be the middle point;

[0109] (3) If the number of adjacent pixels of a pixel is greater than two, then the pixel type is determined as a bifurcation point.

[0110] Here, since a bifurcation point may correspond to more than one branch blood vessel, and each branch blood vessel corresponds to an adjacent pixel, if the number of adjacent pixels of a pixel is greater than two, then the pixel type is determined to be a bifurcation point.

[0111] 4) Among the point types, the pixels belonging to the endpoints and bifurcation points are identified as target pixels;

[0112] 5) Perform depth-first processing on the target pixels to obtain a tree-like topology structure composed of the target pixels;

[0113] 6) Using the anatomical structure of the coronary arteries, the tree-like topology is mapped to obtain the naming of each major branch in the tree-like topology.

[0114] In steps 1)-6) above, the anterior points of the main coronary arteries and major branches are extracted, and these anterior points are skeletonized to obtain the corresponding centerline results. For an example, please refer to [link to example]. Figure 3 , Figure 3 This is a schematic diagram of the centerline result of a major branch provided in an embodiment of this application, as shown below. Figure 3 As shown, according to Figure 3 The centerline results of the main branches shown indicate the key points to be identified on the centerline. Figure 3 The black and white dots in the center line result; specifically, it is necessary to calculate the number of adjacent pixels for each pixel in the center line result. For each pixel, if the number of adjacent pixels is equal to 1, then the pixel is an endpoint; if the number of adjacent pixels is equal to 3, then the pixel is a branch point; if the number of adjacent pixels is equal to 2, then the pixel is a midpoint. The endpoints ( Figure 3 (white dot in the middle) and bifurcation point ( Figure 3 The black dots in the diagram are identified as key points in the centerline result. A tree topology is then built using a depth-first traversal based on these key points. For an example, please refer to [link to example]. Figure 4 , Figure 4 This is a schematic diagram of a topology provided in an embodiment of this application, such as... Figure 4 As shown, Figure 3 The endpoints and bifurcation points in the graph are processed using depth-first search to obtain the graph. Figure 4 The tree-like topology shown includes the first bifurcation point LAD and the second endpoint D1, the second bifurcation point LAD and the third endpoint D2, the third bifurcation point LAD and the fourth endpoint D3, and the fourth bifurcation point LAD and the fifth endpoint D4, all based on the anatomy of the coronary arteries. Figure 4 The tree topology is mapped to obtain the name of each main branch in the tree topology; for an example, please refer to [link to example]. Figure 5 , Figure 5 This is a schematic diagram illustrating the naming of a main branch as provided in an embodiment of this application, such as... Figure 5 As shown, the tree-like topology can determine the order from top to bottom, thus determining the naming of specific branch vessels; the main coronary artery is named LAD, and in clinical practice, based on the anatomical structure, LAD is the main branch, and the branches from top to bottom are named as follows: the first branch is named D1, the second branch is named D2, the third branch is named D3, the fourth branch is named D4, and so on. Here, D is an abbreviation for diagonal branch.

[0115] S1054. The naming of the precursor points of the minor branch categories, the naming of the precursor points of the major coronary artery categories, and the naming of each major branch are determined as the naming result of the blood vessel.

[0116] In summary, this embodiment employs a vascular classification model to categorize vascular images, then trains corresponding point cloud segmentation models for different vascular image categories. Furthermore, only the foreground information of the image is input into the point cloud segmentation model to avoid noise interference, thereby improving the accuracy of fine classification for different vascular categories. Additionally, the coordinate features in the foreground information preserve the image's topological structure, allowing the trained point cloud segmentation model to predict secondary branches and major coronary arteries. Then, the topological structure and vascular anatomy are used to further determine the naming of each vessel in the major branches, thus improving the accuracy of vascular naming. The classification-then-segmentation, coarse-to-fine naming technique provided in this embodiment enables more accurate vascular naming.

[0117] This application provides a method for naming blood vessels, comprising: acquiring an angiographic image of a blood vessel; inputting the angiographic image into a pre-trained blood vessel classification model to obtain a blood vessel category corresponding to the angiographic image; inputting the angiographic image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiographic image, and extracting point cloud features from the blood vessel segmentation image to obtain point cloud features of foreground points in the blood vessel segmentation image; based on the blood vessel category corresponding to the angiographic image, inputting the point cloud features of foreground points in the blood vessel segmentation image into a pre-trained point cloud segmentation model corresponding to the blood vessel category to obtain a fine classification result for each foreground point in the blood vessel segmentation image; establishing the topological structure of the blood vessel based on the fine classification result of each foreground point in the blood vessel segmentation image, and obtaining the naming result of the blood vessel based on the topological structure of the blood vessel.

[0118] Thus, the technical solution provided in this application can classify blood vessels in angiographic images, train corresponding point cloud segmentation models for different categories of angiographic images, and input images into the point cloud segmentation model that only have foreground information, which can avoid noise interference, thereby improving the accuracy of the fine classification results for different categories of blood vessels. Furthermore, the topological structure of the image is preserved through point cloud features, and finally, the topological structure is combined to obtain the final blood vessel naming result, which improves the accuracy of blood vessel naming.

[0119] Based on the same application concept, this application also provides a blood vessel naming device corresponding to the blood vessel naming method provided in the above embodiment. Since the principle of the device in this application is similar to the blood vessel naming method in the above embodiment, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0120] Please see Figure 6 , Figure 7 , Figure 6 This is one of the structural diagrams of a blood vessel naming device provided in an embodiment of this application. Figure 7 This is a second structural diagram of a blood vessel naming device provided in an embodiment of this application. Figure 6 As shown, the naming device 610 includes:

[0121] The acquisition module 611 is used to acquire angiographic images of blood vessels;

[0122] The first processing module 612 is used to input the angiography image into a pre-trained blood vessel classification model to obtain the blood vessel category corresponding to the angiography image;

[0123] The second processing module 613 is used to input the angiography image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiography image, and to extract point cloud features from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image.

[0124] The third processing module 614 is used to input the point cloud features of the foreground points in the blood vessel segmentation image into the pre-trained point cloud segmentation model corresponding to the blood vessel category based on the blood vessel category corresponding to the angiography image, so as to obtain the fine classification result of each foreground point in the blood vessel segmentation image.

[0125] The naming module 615 is used to establish the topological structure of the blood vessels based on the fine classification results of each foreground point in the blood vessel segmentation image, and to obtain the naming result of the blood vessels based on the topological structure of the blood vessels.

[0126] Optionally, when the second processing module 613 is used to extract point cloud features from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image, the second processing module 613 is specifically used for:

[0127] Obtain the predicted label value, location coordinates, and grayscale value of each point cloud in the blood vessel segmentation image;

[0128] Based on the predicted label value of each point cloud, foreground points are extracted from the point cloud of the blood vessel segmentation image.

[0129] The position coordinates of the foreground point in the blood vessel segmentation image are normalized to obtain the normalized coordinates of the foreground point.

[0130] For each foreground point in the blood vessel segmentation image, the position coordinates of the foreground point, the normalized coordinates of the foreground point, and the gray value of the foreground point are concatenated to obtain the point cloud feature of the foreground point in the blood vessel segmentation image.

[0131] Optional, such as Figure 7 As shown, the naming device 610 further includes a training module 616, which is used for:

[0132] Obtain the label of each foreground point in the blood vessel segmentation image;

[0133] For each foreground point in a sample, the point cloud features of the foreground point in the sample are concatenated with the label of the foreground point in the sample to obtain the target features of the foreground point.

[0134] For each foreground point, the target features of that foreground point are input into the point cloud segmentation model corresponding to the blood vessel category to obtain the predicted fine classification result of that foreground point;

[0135] For each foreground point, the predicted sub-classification result of the foreground point and the label of the foreground point are used to calculate the loss value. If the loss value is not less than a preset threshold, the model parameters of the point cloud segmentation model are adjusted to continue training the point cloud segmentation model. If the loss value is less than the preset threshold, the training is completed, and the trained point cloud segmentation model corresponding to the blood vessel category is obtained.

[0136] Optionally, when the naming module 615 is used to establish the topological structure of the blood vessels based on the fine classification results of each foreground point in the blood vessel segmentation image, and to obtain the naming result of the blood vessels based on the topological structure of the blood vessels, the naming module 615 is specifically used for:

[0137] Based on the fine classification results of each foreground point in the blood vessel segmentation image, the foreground point indicated by the fine classification results as a fine branch category is obtained, and the name corresponding to the fine branch category is determined as the name of the foreground point of the fine branch category;

[0138] The leading point of the main coronary artery category is obtained from the detailed classification results, and the name corresponding to the main coronary artery category is determined as the name of the leading point of the main coronary artery category;

[0139] The foreground points indicating the main branch categories are obtained from the fine classification results. Based on the foreground points of the main branch categories, a topology structure composed of the foreground points of the main branch categories is established, and the name of each main branch in the topology structure is obtained.

[0140] The naming of the precursor points of the minor branch categories, the naming of the precursor points of the major coronary artery categories, and the naming of each major branch are determined as the naming results of the blood vessels.

[0141] Optionally, when the naming module 615 is used to establish a topology structure composed of the foregrounds of the main branch categories based on the foregrounds of the main branch categories, and to obtain the name of each main branch in the topology structure, the naming module 615 is specifically used for:

[0142] Based on the anterior points of the major branch categories, the anterior points of the major branch categories and the anterior points of the major coronary artery categories are determined as target anterior points;

[0143] The target foreground point is subjected to skeletonization processing to obtain the centerline results of the main branches of the blood vessel;

[0144] Determine the number of adjacent pixels for each pixel in the centerline result, and determine the pixel type for each pixel based on the number of adjacent pixels.

[0145] In the point type, the pixels belonging to the endpoints and bifurcation points are determined as target pixels;

[0146] The target pixels are subjected to depth-first processing to obtain a tree-like topology structure composed of the target pixels;

[0147] The tree-like topology is mapped using the anatomical structure of the coronary arteries to obtain the names of each major branch in the tree-like topology.

[0148] Optionally, when the naming module 615 is used to determine the point type of each pixel based on the number of adjacent pixels of each pixel, the naming module 615 is specifically used for:

[0149] Based on the number of adjacent pixels of each pixel, if the number of adjacent pixels of each pixel is one, then the pixel type of the pixel is determined to be an endpoint.

[0150] If a pixel is adjacent to two other pixels, then the pixel type is determined to be the middle point.

[0151] If a pixel has more than two adjacent pixels, then the pixel type is determined to be a bifurcation point.

[0152] This application provides a blood vessel naming device, comprising: an acquisition module for acquiring an angiographic image of a blood vessel; a first processing module for inputting the angiographic image into a pre-trained blood vessel classification model to obtain a blood vessel category corresponding to the angiographic image; a second processing module for inputting the angiographic image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiographic image, and extracting point cloud features from the blood vessel segmentation image to obtain point cloud features of foreground points in the blood vessel segmentation image; a third processing module for inputting the point cloud features of foreground points in the blood vessel segmentation image into a pre-trained point cloud segmentation model corresponding to the blood vessel category, based on the blood vessel category corresponding to the angiographic image, to obtain a fine classification result for each foreground point in the blood vessel segmentation image; and a naming module for establishing the topological structure of the blood vessel based on the fine classification result of each foreground point in the blood vessel segmentation image, and obtaining the naming result of the blood vessel based on the topological structure of the blood vessel.

[0153] Thus, the technical solution provided in this application can classify blood vessels in angiographic images, train corresponding point cloud segmentation models for different categories of angiographic images, and input images into the point cloud segmentation model that only have foreground information, which can avoid noise interference, thereby improving the accuracy of the fine classification results for different categories of blood vessels. Furthermore, the topological structure of the image is preserved through point cloud features, and finally, the topological structure is combined to obtain the final blood vessel naming result, which improves the accuracy of blood vessel naming.

[0154] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 8 As shown, the electronic device 800 includes a processor 810, a memory 820, and a bus 830.

[0155] The memory 820 stores machine-readable instructions executable by the processor 810. When the electronic device 800 is running, the processor 810 communicates with the memory 820 via the bus 830. When the machine-readable instructions are executed by the processor 810, they can perform the operations described above. Figure 1 as well as Figure 2 The steps of the blood vessel naming method in the illustrated method embodiment can be found in the method embodiment for specific implementation, and will not be repeated here.

[0156] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can perform the above-described actions. Figure 1 as well as Figure 2The steps of the blood vessel naming method in the illustrated method embodiment can be found in the method embodiment for specific implementation, and will not be repeated here.

[0157] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0158] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0159] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0160] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0161] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0162] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The scope of protection of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method of naming a blood vessel, characterized by, The naming method includes: Obtain angiographic images of blood vessels; The angiography image is input into a pre-trained blood vessel classification model to obtain the blood vessel category corresponding to the angiography image; The angiography image is input into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiography image. Point cloud features are extracted from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image. Based on the blood vessel category corresponding to the angiography image, the point cloud features of the foreground points in the blood vessel segmentation image are input into the pre-trained point cloud segmentation model corresponding to the blood vessel category to obtain the fine classification result of each foreground point in the blood vessel segmentation image; Based on the fine classification results of each foreground point in the blood vessel segmentation image, the topological structure of the blood vessel is established, and based on the topological structure of the blood vessel, the naming result of the blood vessel is obtained. The steps of establishing the topological structure of the blood vessels based on the fine classification results of each foreground point in the segmented blood vessel image, and obtaining the naming results of the blood vessels based on the topological structure of the blood vessels, include: Based on the fine classification results of each foreground point in the blood vessel segmentation image, the foreground point indicated by the fine classification results as a fine branch category is obtained, and the name corresponding to the fine branch category is determined as the name of the foreground point of the fine branch category; The leading point of the main coronary artery category is obtained from the detailed classification results, and the name corresponding to the main coronary artery category is determined as the name of the leading point of the main coronary artery category; The foreground points indicating the main branch categories are obtained from the fine classification results. Based on the foreground points of the main branch categories, a topology structure composed of the foreground points of the main branch categories is established, and the name of each main branch in the topology structure is obtained. The naming of the precursor points of the minor branch categories, the naming of the precursor points of the major coronary artery categories, and the naming of each major branch are determined as the naming results of the blood vessels; The step of establishing a topology structure composed of the foreground points of the main branch categories based on the foreground points of the main branch categories, and obtaining the naming of each main branch in the topology structure, includes: Based on the anterior points of the major branch categories, the anterior points of the major branch categories and the anterior points of the major coronary artery categories are determined as target anterior points; The target foreground point is subjected to skeletonization processing to obtain the centerline results of the main branches of the blood vessel; Determine the number of adjacent pixels for each pixel in the centerline result, and determine the pixel type for each pixel based on the number of adjacent pixels. In the point type, the pixels belonging to the endpoints and bifurcation points are determined as target pixels; The target pixels are subjected to depth-first processing to obtain a tree-like topology structure composed of the target pixels; The tree-like topology is mapped using the anatomical structure of the coronary arteries to obtain the names of each major branch in the tree-like topology.

2. The naming method of claim 1, wherein, The following steps are used to extract point cloud features from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image: Obtain the predicted label value, location coordinates, and grayscale value of each point cloud in the blood vessel segmentation image; Based on the predicted label value of each point cloud, foreground points are extracted from the point cloud of the blood vessel segmentation image. The position coordinates of the foreground point in the blood vessel segmentation image are normalized to obtain the normalized coordinates of the foreground point. For each foreground point in the blood vessel segmentation image, the position coordinates of the foreground point, the normalized coordinates of the foreground point, and the gray value of the foreground point are concatenated to obtain the point cloud feature of the foreground point in the blood vessel segmentation image.

3. The naming method of claim 1, wherein, The point cloud segmentation model is trained using the following steps: Obtain the label of each foreground point in the blood vessel segmentation image; For each foreground point in a sample, the point cloud features of the foreground point in the sample are concatenated with the label of the foreground point in the sample to obtain the target features of the foreground point. For each foreground point, the target features of the foreground point are input into the point cloud segmentation model corresponding to the blood vessel category to obtain the predicted fine classification result of the foreground point; For each foreground point, the predicted sub-classification result of the foreground point and the label of the foreground point are used to calculate the loss value. If the loss value is not less than a preset threshold, the model parameters of the point cloud segmentation model are adjusted to continue training the point cloud segmentation model. If the loss value is less than the preset threshold, the training is completed, and the trained point cloud segmentation model corresponding to the blood vessel category is obtained.

4. The naming method of claim 1, wherein, The step of determining the pixel type of each pixel based on the number of adjacent pixels of each pixel includes: Based on the number of adjacent pixels of each pixel, if the number of adjacent pixels of each pixel is one, then the pixel type of the pixel is determined to be an endpoint. If a pixel is adjacent to two other pixels, then the pixel type is determined to be the middle point. If a pixel has more than two adjacent pixels, then the pixel type is determined to be a bifurcation point.

5. A device for naming blood vessels, characterized in that, The naming device includes: The acquisition module is used to acquire angiographic images of blood vessels; The first processing module is used to input the angiography image into a pre-trained blood vessel classification model to obtain the blood vessel category corresponding to the angiography image; The second processing module is used to input the angiography image into a pre-trained blood vessel segmentation model to obtain a blood vessel segmentation image corresponding to the angiography image, and to extract point cloud features from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image. The third processing module is used to input the point cloud features of the foreground points in the blood vessel segmentation image into the pre-trained point cloud segmentation model corresponding to the blood vessel category based on the blood vessel category of the angiography image, so as to obtain the fine classification result of each foreground point in the blood vessel segmentation image. The naming module is used to establish the topological structure of the blood vessels based on the fine classification results of each foreground point in the blood vessel segmentation image, and to obtain the naming result of the blood vessels based on the topological structure of the blood vessels. When the naming module is used to establish the topological structure of the blood vessels based on the fine classification results of each foreground point in the blood vessel segmentation image, and to obtain the naming result of the blood vessels based on the topological structure of the blood vessels, the naming module is specifically used for: Based on the fine classification results of each foreground point in the blood vessel segmentation image, the foreground point indicated by the fine classification results as a fine branch category is obtained, and the name corresponding to the fine branch category is determined as the name of the foreground point of the fine branch category; The leading point of the main coronary artery category is obtained from the detailed classification results, and the name corresponding to the main coronary artery category is determined as the name of the leading point of the main coronary artery category; The foreground points indicating the main branch categories are obtained from the fine classification results. Based on the foreground points of the main branch categories, a topology structure composed of the foreground points of the main branch categories is established, and the name of each main branch in the topology structure is obtained. The naming of the precursor points of the minor branch categories, the naming of the precursor points of the major coronary artery categories, and the naming of each major branch are determined as the naming results of the blood vessels; When the naming module is used to establish a topology structure composed of the foreground points of the main branch categories based on the foreground points of the main branch categories, and to obtain the name of each main branch in the topology structure, the naming module is specifically used for: Based on the anterior points of the major branch categories, the anterior points of the major branch categories and the anterior points of the major coronary artery categories are determined as target anterior points; The target foreground point is subjected to skeletonization processing to obtain the centerline results of the main branches of the blood vessel; Determine the number of adjacent pixels for each pixel in the centerline result, and determine the pixel type for each pixel based on the number of adjacent pixels. In the point type, the pixels belonging to the endpoints and bifurcation points are determined as target pixels; The target pixels are subjected to depth-first processing to obtain a tree-like topology structure composed of the target pixels; The tree-like topology is mapped using the anatomical structure of the coronary arteries to obtain the names of each major branch in the tree-like topology.

6. The naming device according to claim 5, characterized in that, When the second processing module extracts point cloud features from the blood vessel segmentation image to obtain the point cloud features of the foreground points in the blood vessel segmentation image, the second processing module is specifically used for: Obtain the predicted label value, location coordinates, and grayscale value of each point cloud in the blood vessel segmentation image; Based on the predicted label value of each point cloud, foreground points are extracted from the point cloud of the blood vessel segmentation image. The position coordinates of the foreground point in the blood vessel segmentation image are normalized to obtain the normalized coordinates of the foreground point. For each foreground point in the blood vessel segmentation image, the position coordinates of the foreground point, the normalized coordinates of the foreground point, and the gray value of the foreground point are concatenated to obtain the point cloud feature of the foreground point in the blood vessel segmentation image.

7. An electronic device, characterized in that, include: The device includes a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and the machine-readable instructions are executed by the processor to perform the steps of the blood vessel naming method as described in any one of claims 1 to 4.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the blood vessel naming method as described in any one of claims 1 to 4.

Images