A vascular matching method, device, equipment and storage medium based on a topological graph

Abstract

The present invention relates to the technical field of blood vessel detection, and specifically relates to a blood vessel matching method, device, equipment and storage medium based on a topological graph. According to the curve information in the topological graph, the present invention classifies curves with different information in the same topological graph, and layers the topological graph according to the classification results, that is, the curves on the same layer have the same information, and the curves on different layers have different information. When performing curve matching, the curves on the corresponding topological layered graph are matched, and the matching of the curves is the matching of the blood vessels corresponding to the curves. Laying the topological graph by the present invention can not only improve the blood vessel matching speed, but also reduce the degree of incorrect matching of blood vessels. Only by correctly matching the blood vessels can it help doctors view the changes of the same blood vessel for the diagnosis of patients.

Description

Technical Field

[0001] The present invention relates to the technical field of blood vessel detection, and in particular to a blood vessel matching method, device, equipment and storage medium based on a topological map. Background Art

[0002] The rapid growth of diabetes has attracted the attention of all walks of life. The exponential growth of diseases caused by diabetes has become a huge challenge facing the current healthcare industry. The number of patients suffering from diseases caused by diabetes continues to grow at an alarming rate. From a medical point of view, diabetes is considered to be the basis of many health problems and later disorders, that is, diabetes can cause a series of lesions and complications, including: causing serious heart disease, diabetic retinopathy (DR) and kidney problems. DR is one of the most common complications of diabetes and is considered one of the main causes of blindness.

[0003] Diabetic retinopathy (DR) is one of the most common microvascular complications of diabetes and one of the main causes of irreversible blindness. The severity of DR can be divided into 0-4 levels: normal, mild, moderate, severe and proliferative DR (PDR). The retinal vascular system is mainly composed of arterioles and venules, and its changes may be early indicators of diabetes-related microvascular damage. With the development of computer image processing technology, computer-assisted analysis programs have been developed to quantitatively evaluate the geometric network parameters of retinal vessels (such as fractal dimension, tortuosity, branching angle, etc.). However, due to the complexity of fundus vascular morphology, the relationship between vascular geometry and DR has not been determined. Therefore, an analysis framework that can objectively and reasonably reflect the geometry of retinal vessels remains to be explored. The existing technology only analyzes the geometric parameters of retinal vessels (such as fractal dimension, tortuosity, branching angle, etc.), but these parameters are contradictory in different analysis works, which reduces the matching efficiency, and therefore cannot objectively reflect the changes in vascular geometry, and thus cannot objectively reflect the characteristics of vascular shape.

[0004] In summary, the existing technology cannot objectively reflect the shape characteristics of blood vessels.

[0005] Therefore, the prior art still needs to be improved and enhanced. Summary of the invention

[0006] To solve the above technical problems, the present invention provides a topological map-based vascular matching method, device, equipment and storage medium, innovatively proposes a model for objectively analyzing vascular shape, and solves the problem of low vascular matching efficiency.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] In a first aspect, the present invention provides a blood vessel matching method based on a topological graph, which includes:

[0009] Obtain each topological graph corresponding to the blood vessels;

[0010] According to the curve information in each of the topological graphs, layer each of the topological graphs to obtain a topological layer graph corresponding to each of the topological graphs, where the curve information is used to reflect the information of the blood vessels;

[0011] Match the curves in each of the topological layer graphs to obtain a matching curve, where the matching curve is used to characterize the matching blood vessels.

[0012] In one implementation, the step of, according to the curve information in each of the topological graphs, layer each of the topological graphs to obtain a topological layer graph corresponding to each of the topological graphs, where the curve information is used to reflect the information of the blood vessels, includes:

[0013] According to the curve information, obtain the curve size information in the curve information, where the curve size information is used to reflect the size information of the blood vessels;

[0014] According to the curve size information, layer each of the topological graphs to obtain a topological layer graph corresponding to each of the topological graphs.

[0015] In one implementation, the step of, according to the curve size information, layer each of the topological graphs to obtain a topological layer graph corresponding to each of the topological graphs, includes:

[0016] According to the curve size information, obtain a first curve size and a second curve size in the curve size information, where the first curve size is different from the second curve size;

[0017] According to the first curve size and the second curve size, layer each of the topological graphs to obtain a topological first layer graph and a topological second layer graph in the topological layer graph, where the topological first layer graph corresponds to the first curve size, and the topological second layer graph corresponds to the second curve size.

[0018] In one implementation, the step of matching the curves in each of the topological layer graphs to obtain a matching curve, where the matching curve is used to characterize the matching blood vessels, includes:

[0019] According to the topological layer graph, obtain the vertices on the curves in the topological layer graph;

[0020] Match the curves according to the vertices to obtain a matching curve.

[0021] In one implementation, the process of obtaining the topological hierarchical graphs corresponding to the respective topological graphs according to the curve dimension information includes:

[0022] Obtaining the degree of curve bending in the curve information according to the curve information;

[0023] Dividing the curves included in the respective topological graphs according to the curve dimension information and the degree of curve bending to obtain a division result;

[0024] Layering the respective topological graphs according to the division result to obtain the topological hierarchical graphs corresponding to the respective topological graphs.

[0025] In one implementation, the process of obtaining the respective topological graphs corresponding to the blood vessel includes:

[0026] Obtaining the original image corresponding to the blood vessel;

[0027] Applying a binarization algorithm to the original image to obtain a binarized image;

[0028] Applying a skeletonization algorithm to the binarized image to obtain a skeletonized image;

[0029] Obtaining the nodes and curves corresponding to the skeletonized image according to the skeletonized image;

[0030] Obtaining the respective topological graphs according to the nodes and the curves.

[0031] In one implementation, it further includes:

[0032] Obtaining a first topological graph and a second topological graph in the topological graph according to the topological graph;

[0033] Obtaining a first topological hierarchical graph corresponding to the first topological graph according to the first topological graph;

[0034] Obtaining a second topological hierarchical graph corresponding to the second topological graph according to the second topological graph;

[0035] Aligning and splicing the curves where the first topological hierarchical graph and the second topological hierarchical graph match to obtain a topologically hierarchical matching graph;

[0036] Obtaining a topological change graph corresponding to the change from the first topological graph to the second topological graph according to the topologically hierarchical matching graph and the first topological graph.

[0037] In a second aspect, an embodiment of the present invention further provides a device for a blood vessel matching method based on topological graphs, where the device includes the following components:

[0038] A topology graph generation module, configured to obtain each topology graph corresponding to blood vessels;

[0039] A topology graph layering module, configured to layer each of the topology graphs according to curve information in each of the topology graphs, so as to obtain a topology layered graph corresponding to each of the topology graphs, where the curve information is used to reflect information of the blood vessels;

[0040] A blood vessel matching module, configured to match curves in each of the topology layered graphs to obtain matching curves, where the matching curves are used to represent matching blood vessels.

[0041] In a third aspect, an embodiment of the present invention further provides a terminal device, where the terminal device includes a memory, a processor, and a topology graph-based blood vessel matching program stored in the memory and executable on the processor. When the processor executes the topology graph-based blood vessel matching program, the steps of the above-mentioned topology graph-based blood vessel matching method are implemented.

[0042] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a topology graph-based blood vessel matching program is stored. When the topology graph-based blood vessel matching program is executed by a processor, the steps of the above-mentioned topology graph-based blood vessel matching method are implemented.

[0043] Beneficial effects: According to the curve information in the topology graph, the present invention classifies curves with different information in the same topology graph, and layers the topology graph according to the classification result, that is, curves on the same layer have the same information, and curves on different layers have different information. When performing curve matching, the curves on the corresponding topology layered graphs are matched, and the matching of curves is the matching of blood vessels corresponding to the curves. Laying the topology graph in the present invention can not only improve the blood vessel matching speed, but also reduce the degree of incorrect matching of blood vessels. Only by correctly matching blood vessels can it help doctors view the changes of the same blood vessel, so as to diagnose patients. Description of the Drawings

[0044] Figure 1 is the overall flowchart of the present invention;

[0045] Figure 2 is the binary image of the present invention;

[0046] Figure 3 is the skeletonized image of the present invention;

[0047] Figure 4 are the nodes and curves in the topology graph of the embodiment;

[0048] Figure 5 are the retinal blood vessels in the embodiment;

[0049] Figure 6 is the main blood vessel in the embodiment;

[0050] Figure 7 is the topological graph of the main blood vessel in the embodiment;

[0051] Figure 8 is the topological graph G corresponding to the retinal blood vessels collected for the first time in the embodiment a ;

[0052] Figures 9 - 11 is the topological change graph in the embodiment;

[0053] Figure 12 is the topological graph G corresponding to the retinal blood vessels collected for the second time in the embodiment b ;

[0054] Figure 13 is the internal structure principle block diagram of the terminal device provided by the embodiment of the present invention. Specific Embodiments

[0055] The technical solutions in the present invention will be clearly and completely described below in conjunction with the embodiments and the accompanying drawings of the specification. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0056] According to research, the rapid growth of diabetes has attracted attention from all walks of life, and the exponential growth of diseases caused by diabetes has become a huge challenge facing the current healthcare industry. The number of patients suffering from diseases caused by diabetes continues to grow at an alarming rate. From a medical point of view, diabetes is considered to be the basis of many health problems and later disorders, that is, diabetes can cause a series of lesions and complications, including: causing serious heart disease, diabetic retinopathy (DR) and kidney problems. DR is one of the most common complications of diabetes and is considered to be one of the main causes of blindness. Diabetic retinopathy (DR) is one of the most common microvascular complications of diabetes and is one of the main causes of irreversible blindness. The severity of DR can be divided into 0-4 levels: normal, mild, moderate, severe and proliferative DR (PDR). The retinal vascular system, mainly composed of arterioles and venules, may be an early indicator of diabetes-related microvascular damage. With the development of computer image processing technology, computer-assisted analysis programs have been developed to quantitatively evaluate the geometric network parameters of retinal blood vessels (such as fractal dimension, tortuosity, branching angle, etc.). However, due to the complexity of fundus vascular morphology, the relationship between vascular geometry and DR has not been determined. Therefore, an analytical framework that can objectively and reasonably reflect the geometric shape of retinal blood vessels remains to be explored. Existing technologies only analyze the geometric parameters of retinal blood vessels (such as fractal dimension, curvature, branching angle, etc.), but these parameters are contradictory in different analysis works, so they cannot objectively reflect the changes in blood vessel geometry, and thus cannot objectively reflect the characteristics of blood vessel shape.

[0057] In order to solve the above technical problems, the present invention provides a method, device, equipment and storage medium for blood vessel matching based on a topological map, innovatively proposes a model for objectively analyzing the shape of blood vessels, and solves the problem of low efficiency of blood vessel matching. In specific implementation, the topological map corresponding to the blood vessel is first collected, and then the topological map is layered according to the curve information in the topological map to obtain a topological layered map, and finally the curve matching is performed according to the topological layered map to achieve blood vessel matching. The present invention layers the topological map to not only increase the blood vessel matching speed, but also reduce the degree of incorrect matching of blood vessels. Only by correctly matching the blood vessels can doctors view the changes in the same blood vessel in order to diagnose the patient.

[0058] For example, in January, the retinal blood vessels of a patient are sampled to obtain the first topological map corresponding to the retinal blood vessels. In October, the retinal blood vessels of the patient are sampled again to obtain the second topological map corresponding to the retinal blood vessels. The topological maps are stratified according to the curve sizes (curve information) in the topological maps. The curves with larger sizes in the first topological map are divided into one layer, and the curves with smaller sizes are divided into another layer. That is, the first topological map is divided into topological stratified maps a1 and a2, where the curve size range s1 in the topological stratified map a1 is greater than the curve size range s2 in the topological stratified map a2. The above method is also used to stratify the second topological map, and the second topological map is divided into topological stratified maps b1 and b2, where the curve size of the topological stratified map b1 is also s1, and the curve size of the topological stratified map b2 is also s2. That is, the curve sizes in the topological stratified map b1 are the same as those in the topological stratified map a1, and the curve sizes in the topological stratified map b2 are the same as those in the topological stratified map a2. When it is necessary to match the curve c1 (c1 is located in the first topological map) corresponding to the blood vessel c and the curve c2 (c2 is located in the second topological map), according to the fact that c1 is within the range s1, as long as the corresponding curve c1 is found in the topological stratified maps a1 and b1, and the change of the curve c1 from the topological stratified map a1 to the topological stratified map b1 is compared, it is possible to know the change of the blood vessel corresponding to the curve c1 from January to October, so as to facilitate the doctor to treat the patient according to the change. In this embodiment, since the topological map is stratified, that is, the curves are divided according to the curve sizes of the topological map, so as long as the curve c1 is searched in the corresponding topological stratified map, the matching speed can be improved.

[0059] Exemplary method

[0060] The blood vessel matching method based on the topological map in this embodiment can be applied to a terminal device, and the terminal device can be a terminal product with an image acquisition function. In this embodiment, as Figure 1 shown, the blood vessel matching method based on the topological map specifically includes the following steps:

[0061] S100, obtain each topological map corresponding to the blood vessel.

[0062] When collecting the image of the blood vessel, due to the presence of other tissues of the human body, it is impossible to accurately distinguish the size and shape of the blood vessel through the image, and then the doctor cannot diagnose the patient. Therefore, in this embodiment, the collected image is first processed to obtain a topological map. The topological map obtained in step S100 includes the following steps S101, S102, S103, S104, S105:

[0063] S101, obtain the original image corresponding to the blood vessel.

[0064] S102. Apply a binarization algorithm to the original image to obtain a binarized image.

[0065] The binarized image is the image as shown in Figure 2 . In the image, there are only white and black, and Figure 2 the white in it is the blood vessel.

[0066] S103. Apply a skeletonization algorithm to the binarized image to obtain a skeletonized image.

[0067] The skeletonized image is the image in which there are only the curves corresponding to the blood vessels and no other human tissues except the blood vessels. In this embodiment, mathematical morphology operations are used to skeletonize the binary image of the manually segmented blood vessel tree to obtain the skeletonized image as shown in Figure 3 .

[0068] S104. According to the skeletonized image, obtain the nodes and curves corresponding to the skeletonized image as shown in Figure 4 .

[0069] Step S104 includes two parts: node detection and curve extraction.

[0070] Node detection: By detecting the number of neighbors of each pixel point, extract the intersection points (pixel points with more than 2 neighbor pixels: more than two neighbors) and end points (pixel points with a degree of 1: one neighbor). The intersection points are the intersections of two curves, and the end points are the vertices of one curve. The end points and intersection points form a set of nodes V:

[0071] V = [v1, v2,..., v N ∈ R N

[0072] Curve extraction: Traverse each node in V. Starting from one node, find the neighbor pixels with a degree of 2 (having two neighbors) until another node is found within the 8-neighborhood (within 8 pixel points). All the curves generated by traversing the nodes are recorded in an adjacency matrix.

[0073] S105. According to the nodes and the curves, obtain each of the topological graphs.

[0074] Represent the topological graph with an adjacency matrix G:

[0075] G = {l ij} ∈ R N×N

[0076] where l ij = l(v i , v j ), and l ij represents the connection between node v i and v jThe curve between them, using 0 and 1 to represent any two nodes v i and v j Whether there is a curve between them, 0 means no curve, 1 means there is a curve. When the number of connecting curves is 1, the element corresponding to G is l ij ; when the number of connecting curves is 0, the element corresponding to G is 0. Since the topological graph is an undirected graph, so there is l(v i , v j ) = l(v j , v i ), G is a symmetric matrix, and the diagonal elements of G are all 0.

[0077] S200, according to the curve information in each of the topological graphs, layer each of the topological graphs to obtain the topological layer graph corresponding to each of the topological graphs, and the curve information is used to reflect the information of the blood vessels.

[0078] The blood vessels targeted in this embodiment are the retinal blood vessels as shown in Figure 5 . The topological graph corresponding to the retinal blood vessels is an extremely complex curve graph. In order to reduce the complexity of subsequent shape analysis, in this embodiment, according to the shape characteristics of the blood vessels, the retinal blood vessel topological structure is divided into two parts, the main blood vessel layer and the fine blood vessel layer, and calculated in parallel according to the diameter size of the blood vessels. The wide line extraction (WLD) algorithm is used to extract the main blood vessels as shown in Figure 6 . This method uses the non-linear filter WLD for line detection to extract the retinal main blood vessels. After obtaining the retinal main blood vessels, the retinal main blood vessels are then topologized to obtain the main blood vessel topological graph as shown in Figure 7 . The main idea of WLD is to measure the similarity between each pixel and its neighborhood, and select the pixels with low similarity as the pixels on the wide line for the detection of wide lines in the blood vessel graph. The detected wide lines are used as the main blood vessel layer, and the rest are used as the fine blood vessel layer.

[0079] Step S200 includes the following steps S201, S202, S203:

[0080] S201, according to the curve information, obtain the curve size information in the curve information, and the curve size information is used to reflect the size information of the blood vessels.

[0081] The curve size information in this embodiment is the thickness of the curve, and the thickness of the curve is the thickness of the blood vessel. The thick curve corresponds to the main blood vessel, and the thin curve corresponds to the fine blood vessel.

[0082] S202, according to the curve size information, obtain the first curve size and the second curve size in the curve size information, and the first curve size is different from the second curve size.

[0083] In this embodiment, the first curve dimension is the dimension corresponding to the main blood vessel, and the second curve dimension is the dimension corresponding to the micro blood vessel.

[0084] S203. According to the first curve dimension and the second curve dimension, layer each of the topological graphs to obtain a topological first layer graph and a topological second layer graph in the topological layer graph. The topological first layer graph corresponds to the first curve dimension, and the topological second layer graph corresponds to the second curve dimension.

[0085] In this embodiment, the curve corresponding to the main blood vessel in the topological graph is divided into one layer, and the curve corresponding to the micro blood vessel is divided into another layer. So that the sizes of all curves in the topological first layer graph are very close, and the curves in the topological second layer graph are also very close.

[0086] In this embodiment, in addition to layering the topological graph by the curve dimension, the topological graph can also be layered according to the bending degree of the curve. The curves with similar bending degrees are divided into the same layer, which can also improve the efficiency of curve matching. When the topological graph is layered according to the bending degree of the curve, step S200 includes the following steps S204, S205, and S206:

[0087] S204. According to the curve information, obtain the bending degree of the curve in the curve information.

[0088] S205. According to the curve dimension information and the bending degree of the curve, divide the curves included in each of the topological graphs to obtain a division result.

[0089] The division result in this embodiment includes two types. One is the curve with a larger bending degree, and the other is the curve with a smaller bending degree or even tending to be a straight line.

[0090] S206. According to the division result, layer each of the topological graphs to obtain the topological layer graph corresponding to each of the topological graphs.

[0091] S300. Match the curves in each of the topological layer graphs to obtain a matching curve, and the matching curve is used to represent the matching blood vessel.

[0092] First, find the vertices of each curve in the topological layer graph, match the vertices in different topological layer graphs, and then the curves in different topological layer graphs are matched.

[0093] The matching of the curves in this embodiment is based on the following principle:

[0094] Since for the same blood vessel topology graph, different orders are adopted to number the nodes, the forms of its adjacency matrices are different. All the analysis and calculations of the blood vessel topology graph are based on the adjacency matrix G. To ensure that the analysis results are not affected by the form of the adjacency matrix and are meaningful, it is necessary to register the vertices of the blood vessel topology graph. The registration of the vertices of a graph, that is, the matching of the graph, is a basic problem in computer science, aiming to achieve the best association of the vertices and edges of two graphs. When the elements in the adjacency matrix G are scalars, the graph matching problem can be described as This is the classic QAP (quadratic assignment programming) problem, where P is a permutation matrix with only one 1 in each row and each column and the rest of the elements are 0, used to rearrange the adjacency matrix G. The well-known factorized graph matching algorithm is adopted to solve the vertex registration problem of the retinal blood vessel graph. When the vertices of two graphs are registered, the pairing relationship between the edges is also determined accordingly.

[0095] In this embodiment, the topology graph adopted can not only realize the matching of blood vessels, but also utilize the limited topology graph to expand the topology graph to obtain more topology graphs, so that doctors can see the change process of blood vessels according to more topology graphs. The expansion of the topology graph in this embodiment includes the following steps S401, S402, S403, S404, S405:

[0096] S401: Obtain a first topology graph and a second topology graph in the topology graph according to the topology graph.

[0097] The final effect to be achieved in this embodiment is to obtain a topology change graph between the first topology graph and the second topology graph.

[0098] S402: Obtain a first topology hierarchical graph corresponding to the first topology graph according to the first topology graph.

[0099] S403: Obtain a second topology hierarchical graph corresponding to the second topology graph according to the second topology graph.

[0100] S404: Align and splice the curves where the first topology hierarchical graph and the second topology hierarchical graph match to obtain a topology hierarchical matching graph.

[0101] The topology hierarchical matching graph in this embodiment is a topology graph obtained by combining the first topology hierarchical graph and the second topology hierarchical graph into one graph.

[0102] S405: Obtain a topology change graph corresponding to the change from the first topology graph to the second topology graph according to the topology hierarchical matching graph and the first topology graph.

[0103] The principles underlying steps S401 - S405 are as follows:

[0104] To observe the process of blood vessel shape changes between different degrees of DR lesions, a geodesic tool is used to display the shape change process between the registered blood vessel topologies. Given a first - layer topology graph and a second - layer topology graph, the nodes of the first - layer topology graph are aligned with the nodes of the second - layer topology graph to obtain a topology - layer matching graph. The calculation formula for the geodesic between the first - layer topology graph and the second - layer topology graph is as follows:

[0105]

[0106] The value range of t is [0, 1]. For each value of t, a topology change graph can be obtained.

[0107] For example, as Figure 8 shows the topology graph G corresponding to the retina blood vessels collected for the first time a , and as Figure 12 shows the topology graph G corresponding to the retina blood vessels collected for the second time b . The topology graph G b is stratified to obtain a first - layer topology graph and a second - layer topology graph. Aligning the curves of the first - layer topology graph and the second - layer topology graph gives a topology - layer matching graph. Through steps S401, S402, S403, S404, and S405 of this embodiment, the topology graphs from G a to G b can be obtained, such as the topology graphs shown in Figure 9 , 10 , and 11. Figure 9 , 10 , and the topology graphs shown in 11 can clearly reflect the process of blood vessel shape changes in the patient's retina during this period, providing objective blood vessel shape analysis data for ophthalmologists to analyze the patient's DR disease course.

[0108] In summary, based on the curve information in the topology graph, the curves with different information in the same topology graph are classified, and the topology graph is stratified according to the classification results, that is, the curves on the same layer have the same information, and the curves on different layers have different information. When performing curve matching, the curves on the corresponding topology - layer graphs are matched. The matching of curves is the matching of the blood vessels corresponding to the curves. Stratifying the topology graph of the present invention can not only improve the blood vessel matching speed but also reduce the degree of incorrect blood vessel matching. Only by correctly matching the blood vessels can doctors view the changes of the same blood vessel to diagnose the patient.

[0109] In addition, the present invention innovatively applies the graph matching algorithm to the problem of fundus vascular shape analysis, which can realize the shape registration of complex curves such as vascular shapes, providing a basis for the analysis of different retinal vascular shapes. By using the difference in blood vessel diameter, the blood vessels are hierarchically processed, and the shapes of blood vessels in different layers in two blood vessel graphs are paired and calculated separately, which can greatly reduce the computational amount. Parallel computing of multiple layers simultaneously can save computing resources. At the same time, the main blood vessels, microvessels, etc. are hierarchically processed, and the changes of blood vessels in different parts under different lesion grades are analyzed specifically, which is more helpful for doctors' analysis and judgment. Provide doctors with a visual process of blood vessel shape changes to facilitate the study of the DR lesion process.

[0110] Exemplary device

[0111] This embodiment also provides a device for a blood vessel matching method based on a topological graph, and the device includes the following components:

[0112] A topological graph generation module, configured to obtain each topological graph corresponding to the blood vessels;

[0113] A topological graph layering module, configured to layer each of the topological graphs according to the curve information in each of the topological graphs to obtain a topological layered graph corresponding to each of the topological graphs, where the curve information is used to reflect the information of the blood vessels;

[0114] A blood vessel matching module, configured to match the curves in each of the topological layered graphs to obtain matching curves, where the matching curves are used to characterize the matching blood vessels.

[0115] Based on the above embodiments, the present invention also provides a terminal device, and its principle block diagram can be as Figure 13 shown. The terminal device includes a processor, a memory, a network interface, a display screen, and a temperature sensor connected through a system bus. Among them, the processor of the terminal device is used to provide computing and control capabilities. The memory of the terminal device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the terminal device is used to communicate with an external terminal through a network connection. The computer program, when executed by the processor, implements a blood vessel matching method based on a topological graph. The display screen of the terminal device can be a liquid crystal display screen or an electronic ink display screen, and the temperature sensor of the terminal device is pre-set inside the terminal device to detect the operating temperature of the internal device.

[0116] Those skilled in the art can understand, Figure 13The block diagram of the principle shown is only the block diagram of the partial structure related to the solution of the present invention, and does not constitute a limitation on the terminal device to which the solution of the present invention is applied. The specific terminal device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0117] In one embodiment, a terminal device is provided. The terminal device includes a memory, a processor, and a topology-based blood vessel matching program stored in the memory and executable on the processor. When the processor executes the topology-based blood vessel matching program, the following operation instructions are implemented:

[0118] Obtain each topology map corresponding to the blood vessel;

[0119] According to the curve information in each of the topology maps, layer each of the topology maps to obtain a topology layer map corresponding to each of the topology maps, where the curve information is used to reflect the information of the blood vessel;

[0120] Match the curves in each of the topology layer maps to obtain a matching curve, where the matching curve is used to characterize the matching blood vessel.

[0121] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the embodiments of the above methods. Among them, any reference to the memory, storage, database, or other media used in the various embodiments provided by the present invention may include non-volatile and / or volatile memories. Non-volatile memories may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0122] In summary, the present invention discloses a blood vessel matching method, device, equipment and storage medium based on a topological graph. The method includes: obtaining each topological graph corresponding to a blood vessel; according to the curve information in each topological graph, hierarchically dividing each topological graph to obtain a topological hierarchical graph corresponding to each topological graph, where the curve information is used to reflect the information of the blood vessel; matching the curves in each topological hierarchical graph to obtain matching curves, where the matching curves are used to represent matching blood vessels. Hierarchically dividing the topological graph in the present invention can not only improve the blood vessel matching speed, but also reduce the degree of incorrect blood vessel matching. Only by correctly matching the blood vessels can it help doctors view the changes of the same blood vessel for the diagnosis of patients.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A vascular matching method based on a topological graph, characterized in that Including: Obtain each topological map corresponding to the blood vessel, and extract the main blood vessel by using the wide-line extraction method; According to the curve information in each of the topological maps, layer each of the topological maps to obtain a topological layered map corresponding to each of the topological maps, where the curve information is used to reflect the information of the blood vessel; Match the curves in each of the topological layered maps to obtain matching curves, where the matching curves are used to characterize the matching blood vessels and are used to view the changes of the same blood vessel; The step of, according to the curve information in each of the topological maps, layer each of the topological maps to obtain a topological layered map corresponding to each of the topological maps, where the curve information is used to reflect the information of the blood vessel, includes: According to the curve information, obtain the curve size information in the curve information, where the curve size information is used to reflect the size information of the blood vessel; According to the curve size information, obtain a first curve size and a second curve size in the curve size information, where the first curve size is different from the second curve size, the first curve size is the size of the main blood vessel, and the second curve size is the size of the fine blood vessel; According to the first curve size and the second curve size, layer each of the topological maps to obtain a topological first layered map and a topological second layered map in the topological layered map, where the topological first layered map corresponds to the first curve size, and the topological second layered map corresponds to the second curve size; Further including: According to the topological map, obtain a topological first map and a topological second map in the topological map; According to the topological first map, obtain a topological first layered map corresponding to the topological first map; According to the topological second map, obtain a topological second layered map corresponding to the topological second map; Align and splice the curves that match between the topological first layered map and the topological second layered map to obtain a topological layered matching map; According to the topological layered matching map and the topological first map, obtain a topological change map corresponding to the change from the topological first map to the topological second map.

2. The blood vessel matching method based on a topological graph according to claim 1, wherein The step of matching the curves in each of the topological layered maps to obtain matching curves, where the matching curves are used to characterize the matching blood vessels, includes: According to the topological layered map, obtain the vertices on the curves in the topological layered map; Match the curves according to the vertices to obtain matching curves.

3. The vascular matching method based on a topological graph according to claim 1, wherein The step of, according to the curve size information, layer each of the topological maps to obtain a topological layered map corresponding to each of the topological maps, includes: According to the curve information, obtain the degree of curve bending in the curve information; According to the curve size information and the degree of curve bending, divide the curves included in each of the topological maps to obtain a division result; According to the division result, layer each of the topological maps to obtain a topological layered map corresponding to each of the topological maps.

4. The method for vascular matching based on a topological graph according to claim 1, wherein The step of obtaining each topological map corresponding to the blood vessel includes: Obtain the original image corresponding to the blood vessel; Apply a binarization algorithm to the original image to obtain a binarized image; Apply a skeletonization algorithm to the binarized image to obtain a skeletonized image; Based on the skeletonized image, obtain the nodes and curves corresponding to the skeletonized image; Based on the nodes and the curves, obtain each of the topological graphs.

5. An apparatus for a vascular matching method based on a topological graph, characterized in that, The device includes the following components: A topological graph generation module, configured to obtain each topological graph corresponding to blood vessels and extract the main blood vessels using a wide-line extraction method; A topological graph layering module, configured to layer each of the topological graphs according to the curve information in each of the topological graphs to obtain a topological layered graph corresponding to each of the topological graphs, where the curve information is used to reflect the information of the blood vessels; A blood vessel matching module, configured to match the curves in each of the topological layered graphs to obtain matching curves, where the matching curves are used to represent the matching blood vessels and are used to view the changes of the same blood vessel; The step of layer each of the topological graphs according to the curve information in each of the topological graphs to obtain a topological layered graph corresponding to each of the topological graphs, where the curve information is used to reflect the information of the blood vessels, includes: Based on the curve information, obtain the curve size information in the curve information, where the curve size information is used to reflect the size information of the blood vessels; Based on the curve size information, obtain a first curve size and a second curve size in the curve size information, where the first curve size is different from the second curve size, the first curve size is the size of the main blood vessel, and the second curve size is the size of the fine blood vessels; Based on the first curve size and the second curve size, layer each of the topological graphs to obtain a topological first layered graph and a topological second layered graph in the topological layered graph, where the topological first layered graph corresponds to the first curve size, and the topological second layered graph corresponds to the second curve size; It further includes: Based on the topological graph, obtain a topological first graph and a topological second graph in the topological graph; Based on the topological first graph, obtain a topological first layered graph corresponding to the topological first graph; Based on the topological second graph, obtain a topological second layered graph corresponding to the topological second graph; Align and splice the curves in the topological first layered graph that match the curves in the topological second layered graph to obtain a topological layered matching graph; Based on the topological layered matching graph and the topological first graph, obtain a topological change graph corresponding to the change from the topological first graph to the topological second graph.

6. A terminal device, characterized in that, The terminal device includes a memory, a processor, and a topological graph-based blood vessel matching program stored in the memory and executable on the processor. When the processor executes the topological graph-based blood vessel matching program, the steps of the topological graph-based blood vessel matching method according to any one of claims 1-4 are implemented.

7. A computer-readable storage medium, characterized in that, A topological graph-based blood vessel matching program is stored on the computer-readable storage medium. When the topological graph-based blood vessel matching program is executed by a processor, the steps of the topological graph-based blood vessel matching method according to any one of claims 1-4 are implemented.

Images