Coronary artery stenosis rate determination system and storage medium

Abstract

The embodiment of the present invention discloses a coronary artery stenosis rate determination system and storage medium, the system includes a processor, the processor is configured to execute the following method: coarse segmentation of a coronary artery image to obtain a coarse segmentation result of the coronary artery; fine segmentation of the coarse segmentation result based on a trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and filling the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result; inputting the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and filling the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result; and determining the coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result. The problem of long data processing time in the existing coronary artery stenosis rate determination method is solved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of image processing, and in particular to a coronary artery stenosis rate determination system and a storage medium. Background Art

[0002] As a major threat to human health, coronary heart disease has received more and more attention. One of the core tasks in the diagnosis of coronary heart disease is to diagnose coronary artery stenosis on coronary CTA images. Missed diagnosis, misdiagnosis, and misjudgment of stenosis rate will bring greater risks to patients. Doctors routinely judge the stenosis rate by observing the coronary artery on CPR and straightened images to find the location of the stenosis and judge the degree of stenosis, but observing stenosis is time-consuming and laborious.

[0003] In order to assist doctors in diagnosing coronary artery stenosis, the existing technology usually uses a stenosis rate determination model to analyze coronary artery images to determine the patient's coronary artery stenosis rate. Although this method can obtain the coronary artery stenosis rate, it has a large amount of data calculation, high requirements on processor performance, and a long image processing time.

[0004] The inventors of the present application discovered during the process of implementing the embodiments of the application that the prior art method for determining the coronary artery stenosis rate at least has the problem of long data processing time. Summary of the invention

[0005] The embodiments of the present invention provide a coronary artery stenosis rate determination system and a storage medium, which solve the problem that the coronary artery stenosis rate determination method in the prior art has at least a long data processing time.

[0006] In a first aspect, an embodiment of the present invention provides a system for determining a coronary artery stenosis rate, comprising a processor, wherein the processor is configured to execute the following method:

[0007] Performing rough segmentation on the coronary artery image to obtain a rough segmentation result of the coronary artery;

[0008] Performing fine segmentation on the coarse segmentation result based on the trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and filling the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result;

[0009] Inputting the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and filling the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result;

[0010] The coronary artery stenosis rate is determined according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result.

[0011] In a second aspect, an embodiment of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions are used to perform the following steps when executed by a computer processor:

[0012] Performing rough segmentation on the coronary artery image to obtain a rough segmentation result of the coronary artery;

[0013] Performing fine segmentation on the coarse segmentation result based on the trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and filling the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result;

[0014] Inputting the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and filling the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result;

[0015] The coronary artery stenosis rate is determined according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result.

[0016] The technical solution of the coronary artery stenosis rate determination system provided in the embodiment of the present invention is to perform a coarse segmentation on the coronary artery image to obtain a coarse segmentation result of the coronary artery; perform a fine segmentation on the coarse segmentation result based on a trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and fill the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result; input the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and fill the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result; and determine the coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result. The coronary artery is coarsely positioned using the trained coarse segmentation model, and the coronary artery is finely segmented using the trained fine segmentation model. Since the coarse segmentation is sparse, the coronary artery is finely segmented on the basis of the coarse segmentation result, which can improve the coronary artery segmentation efficiency while ensuring the accuracy of the coronary artery segmentation result, thereby improving the speed of determining the coronary artery stenosis rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 is a flow chart of a method for determining coronary artery stenosis rate provided in Embodiment 1 of the present invention;

[0019] Figure 2 This is a flow chart of a method for determining coronary artery stenosis rate provided in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described through implementation methods with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Embodiment 1

[0022] Figure 1 : is a flow chart of a method for determining a coronary artery stenosis rate provided by an embodiment of the present invention. The technical solution of this embodiment is applicable to the case where a patient's coronary artery stenosis rate is determined based on a coronary artery image. The method can be executed by a processor of a coronary artery stenosis rate determination system provided by an embodiment of the present invention. The method specifically comprises the following steps:

[0023] S101 . Perform a rough segmentation on the coronary artery image to obtain a rough segmentation result of the coronary artery.

[0024] Among them, coronary artery generally refers to the coronary artery, which is the artery that supplies blood to the heart.

[0025] The coronary artery image may be selected as a CT angiography (CTA, Computed Tomography angiography) image.

[0026] A coronary artery image of the patient is acquired, the coronary artery image is downsampled to a preset coarse segmentation resolution to obtain a first image, and the first image is input into a trained coarse segmentation model to obtain a coarse segmentation result of the coronary artery; the coarse segmentation result is upsampled to an original resolution to update the coarse segmentation result, wherein the original resolution is the same as the resolution of the coronary artery image.

[0027] Coarse segmentation is used to locate the coronary arteries in the coronary image. The coronary segmentation accuracy requirement is relatively low. Therefore, the coronary artery location speed can be improved by reducing the resolution of the input image of the trained coarse segmentation model. In this embodiment, the preset coarse segmentation resolution can be set according to actual needs, such as according to the processor performance and the expected image processing time. For example, the coronary image resolution is 1024×1024×1024, and the preset coarse segmentation resolution can be set to 192×192×192, or even 128×128×128.

[0028] The coarse segmentation model may be created based on existing algorithms such as U-Net and machine learning, which is not specifically limited in this embodiment.

[0029] In this embodiment, the coarse segmentation result is a binary segmentation result. After the coarse segmentation result is obtained, it is upsampled to the original resolution to update the coarse segmentation result, and then subjected to noise reduction processing to improve its image quality. Among them, the noise reduction method can be selected as follows: calculating the number of pixels of each connected domain in the updated coarse segmentation result, deleting the connected domain whose number of pixels is less than a preset pixel threshold to update the coarse segmentation result again.

[0030] In one embodiment, the updated rough segmentation result is expanded to update the rough segmentation result again. The expansion process ensures that there is no coronary artery breakage, coronary artery incompleteness, or the segmentation edge does not completely surround the coronary artery edge in the final rough segmentation result.

[0031] S102 . Perform fine segmentation on the coarse segmentation result based on the trained fine segmentation model to obtain a fine segmentation result of the coronary artery.

[0032] Fine segmentation is a fine adjustment based on coarse segmentation. Since coarse segmentation is sparse, the fine segmentation model is a sparse model, such as point cloud segmentation model, graph network, sparse Attention, etc.

[0033] The coordinates of each pixel in the coarse segmentation result and the pixel value of each pixel coordinate in the coronary artery image are input into the trained fine segmentation model to obtain the fine segmentation result of the coronary artery. It can be understood that by using the coordinates of each pixel in the coarse segmentation result and the pixel value of each pixel coordinate in the coronary artery image as input data of the trained fine segmentation model, the accuracy of the pixel value of each coordinate in the fine segmentation result can be improved while ensuring the accuracy of the coordinates of the fine segmentation result.

[0034] Since the pixel coordinates of the coarse segmentation result input into the trained fine segmentation model correspond one-to-one with the pixel coordinates of the coronary artery image, this embodiment fills the fine segmentation result back into the coronary artery image to form a segmentation of the same size as the original image to obtain the coronary artery segmentation result.

[0035] In one embodiment, the number of pixels of each connected domain in the coronary segmentation image is calculated, and the connected domain with a pixel number less than a preset pixel threshold is deleted to update the coronary segmentation image.

[0036] S103 , inputting the coronary artery image into the trained plaque detection model to obtain a plaque prediction result, and filling the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result.

[0037] In one embodiment, the coronary artery image is directly input into the trained plaque detection model to obtain the plaque prediction result.

[0038] In one embodiment, the minimum coronary artery circumscribed frame is determined based on the fine segmentation result, and the minimum coronary artery circumscribed frame is expanded by N pixels to obtain the coronary artery circumscribed frame, and the coronary artery regional image is deducted from the coronary artery image using the coronary artery circumscribed frame, and the coronary artery regional image is input into the trained plaque detection model to obtain the plaque prediction result, and then the plaque prediction result is filled back into the coronary artery image to obtain the plaque segmentation result, wherein N is a natural number, such as 10 or 15. It can be understood that determining the input data of the plaque detection model based on the coronary artery circumscribed frame corresponding to the fine segmentation result can reduce the data calculation amount of the trained plaque detection model, which helps to improve the efficiency of plaque detection, thereby reducing the data calculation amount of the coronary artery stenosis rate and improving the speed of determining the coronary artery stenosis rate.

[0039] In one embodiment, in order to further improve the speed of determining the coronary stenosis rate, the deducted coronary regional image is downsampled to a preset plaque resolution to update the coronary regional image, and then the updated coronary regional image is input into the trained plaque detection model to obtain a plaque prediction result. The resolution rate of the plaque prediction result is the same as that of the updated coronary regional image. In this embodiment, the plaque prediction result is upsampled to the original resolution to update the plaque prediction result, and the updated plaque prediction result is filled back into the coronary image to obtain a plaque segmentation result. In this embodiment, the preset plaque resolution can be selected as 128×192×192, or other low-resolution sizes. The plaque detection model can be created based on a U-Net network or a machine learning algorithm.

[0040] In one embodiment, after the plaque prediction result is filled back into the coronary artery image to obtain the plaque segmentation result, the plaque segmentation result is filtered using the fine segmentation result to update the plaque segmentation result. Using the fine segmentation result to filter the plaque segmentation result can effectively filter out the plaque outside the coronary artery. The number of pixels of each connected domain in the updated plaque segmentation result is calculated, and the connected domain with a pixel number less than the set pixel threshold is deleted to update the plaque segmentation result again.

[0041] S104. Determine the coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result.

[0042] The coronary artery stenosis rate is an indicator used to judge coronary artery stenosis. The higher the stenosis rate, the more serious the coronary artery blockage. When the stenosis reaches a certain degree and causes myocardial ischemia, it will lead to coronary heart disease.

[0043] Since the coronary artery stenosis rate is more accurate when viewed on a straightened coronary artery image, this embodiment determines the coronary artery centerline based on the coronary artery segmentation result; the coronary artery is straightened based on the coronary artery centerline for the coronary artery image, the coronary artery segmentation result, and the plaque segmentation result. The position range of each plaque in the coronary artery segmentation result after the coronary artery is straightened is determined, and the stenosis rate of each plaque is calculated; the maximum stenosis rate is used as the patient's coronary artery stenosis rate, and based on the mapping relationship between the straightened coronary artery segmentation result and the coronary artery image, the coronary artery stenosis rate and the plaque position range corresponding to the coronary artery stenosis rate are mapped to the straightened coronary artery image; based on the mapping relationship between the coronary artery image after the coronary artery is straightened and the coronary artery image, the coronary artery stenosis rate and the plaque position range corresponding to the coronary artery stenosis rate are displayed on the coronary artery image.

[0044] In one embodiment, the Centerlines method of VMTK is used to extract the coronary centerline of the coronary segmentation result, and then the coronary image, plaque segmentation result and coronary segmentation result are straightened based on the coronary centerline. VMTK is a collection of image-based libraries and tools that can perform functions such as vascular 3D reconstruction, geometric analysis, network generation and surface data analysis.

[0045] It can be understood that since the resolutions of the plaque segmentation result, the coronary segmentation result and the coronary image are all the original resolution, the three have the same size. The position of the plaque in the plaque segmentation result after the coronary artery is straightened, the coronary segmentation result after the coronary artery is straightened and the coronary image after the coronary artery is straightened is the same, so the position range of the plaque in the coronary segmentation result after the coronary artery is straightened can be directly determined according to the position range of the plaque in the plaque segmentation result after the coronary artery is straightened; the position range of the plaque in the coronary image after the coronary artery is straightened can be directly determined according to the position range of the plaque in the coronary segmentation result after the coronary artery is straightened.

[0046] In one embodiment, the position ranges of all plaques in the coronary segmentation results after the coronary artery is straightened are mapped to the coronary artery image after the coronary artery is straightened, and then the position ranges of all plaques in the coronary artery image after the coronary artery is straightened are mapped to the coronary artery image, and the stenosis rate of the plaque is displayed on each plaque in the coronary artery image. At the same time, the coronary artery stenosis rate and the plaque corresponding to the coronary artery stenosis rate are marked with preset symbols, such as red characters for the coronary artery stenosis rate and yellow characters for the stenosis rates of other plaques; the pixel color of the plaque position range corresponding to the coronary artery stenosis rate is assigned to orange, and the pixel color of other plaque position ranges is assigned to pink.

[0047] The technical solution of the coronary artery stenosis rate determination system provided in the embodiment of the present invention is to perform a coarse segmentation on the coronary artery image to obtain a coarse segmentation result of the coronary artery; perform a fine segmentation on the coarse segmentation result based on a trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and fill the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result; input the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and fill the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result; and determine the coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result. The coronary artery is coarsely positioned using the trained coarse segmentation model, and the coronary artery is finely segmented using the trained fine segmentation model. Since the coarse segmentation is sparse, the coronary artery is finely segmented on the basis of the coarse segmentation result, which can improve the coronary artery segmentation efficiency while ensuring the accuracy of the coronary artery segmentation result, thereby improving the speed of determining the coronary artery stenosis rate.

[0048] Embodiment 2

[0049] Figure 2 Flow chart of the method for determining coronary artery stenosis rate provided by the second embodiment of the present invention. Based on any of the above embodiments, the embodiment of the present invention optimizes the method for determining coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result as follows: the coronary artery stenosis rate is determined according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result based on the area method and / or the diameter method.

[0050] Accordingly, the method for determining the coronary artery stenosis rate of this embodiment includes:

[0051] S201 . Perform a rough segmentation on the coronary artery image to obtain a rough segmentation result of the coronary artery.

[0052] S202 , performing fine segmentation on the coarse segmentation result based on the trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and filling the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result.

[0053] S203. Display the coronary artery stenosis rate and the plaque position range corresponding to the coronary artery stenosis rate on the coronary artery image according to the mapping relationship between the coronary artery image after the coronary artery is straightened and the coronary artery image.

[0054] S204. Determine the coronary artery stenosis rate based on the area method and / or the diameter method according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result.

[0055] When the coronary artery stenosis rate is determined based on the area method, the coronary artery stenosis rate = (a 2 -b 2 ) / a 2 Among them, a is the lumen diameter of the proximal normal coronary artery, and b is the lumen diameter of the coronary artery at its narrowest point.

[0056] When determining the coronary artery stenosis rate based on the diameter method, the coronary artery stenosis rate = (ab) / a, where a is the lumen diameter of the proximal normal coronary artery, and b is the lumen diameter of the coronary artery at its narrowest point.

[0057] In one embodiment, based on a preset interval, the coronary artery lumen diameter in the coronary artery segmentation result after the coronary artery is straightened is obtained to obtain a first diameter set; the position range of each plaque in the plaque segmentation result after the coronary artery is straightened is determined; based on the preset interval and the position range of each plaque in the coronary artery segmentation result after the coronary artery is straightened, the coronary artery lumen diameter in the position range of each plaque is obtained to obtain a second diameter set; based on the area method and / or the diameter method, the lumen diameter at the maximum stenosis of each plaque is determined from the second diameter set to obtain a first lumen diameter subset; the proximal normal lumen diameter corresponding to each plaque is determined from the first diameter set to obtain a second lumen diameter subset; based on the area method and / or the diameter method, the stenosis rate of each plaque is determined according to the first lumen diameter subset and the second lumen diameter subset.

[0058] In one embodiment, the preset interval can be set according to the actual stenosis rate accuracy requirement, such as 0.1 mm, 0.2 mm, and 0.25 mm.

[0059] In one embodiment, a method for determining the proximal normal lumen diameter corresponding to each plaque from a first diameter set includes: determining the position range of each plaque in the coronary segmentation image after the coronary artery is straightened, and taking the coronary lumen diameter of point M between the position range and the heart as the proximal normal lumen diameter corresponding to the corresponding plaque, wherein point M is m pixels away from the proximal heart end of the position range, wherein m is a preset pixel threshold and is a natural number.

[0060] In one embodiment, the obtained first diameter set is filtered to update the first diameter set to prevent the existence of abnormal values. Specifically, the first diameter set is filtered along the extension direction of the coronary artery using Gaussian filtering to update the first diameter set.

[0061] The maximum stenosis rate determined based on the area method is used as the first coronary artery stenosis rate, and the maximum stenosis rate determined based on the diameter method is used as the second coronary artery stenosis rate. In one embodiment, the first coronary artery stenosis rate and the plaque range corresponding to the first coronary artery stenosis rate are displayed on the coronary artery image, and the second coronary artery stenosis rate and the plaque range corresponding to the second coronary artery stenosis rate are displayed on the coronary artery image, and the first coronary artery stenosis rate and the second coronary artery stenosis rate are distinguished by different marks. For example, the first coronary artery stenosis rate and the second coronary artery stenosis rate are distinguished by different color marks.

[0062] The technical solution of the system for determining the coronary artery stenosis rate provided in the embodiment of the present invention determines the coronary artery stenosis rate by using the diameter method and / or the area method, and the calculation method is simple and the calculation result is accurate.

[0063] Embodiment 3

[0064] The embodiment of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions are used to perform the following coronary artery stenosis rate determination method when executed by a computer processor:

[0065] Performing rough segmentation on the coronary artery image to obtain a rough segmentation result of the coronary artery;

[0066] Performing fine segmentation on the coarse segmentation result based on the trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and filling the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result;

[0067] Inputting the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and filling the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result;

[0068] The coronary artery stenosis rate is determined according to the coronary artery image, the plaque segmentation result and the coronary artery segmentation result.

[0069] Of course, the computer executable instructions of a storage medium including computer executable instructions provided in an embodiment of the present invention are not limited to the operations of the method described above, but can also execute related operations in the method for determining the coronary artery stenosis rate provided in any embodiment of the present invention.

[0070] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (FLASH), a hard disk or an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the coronary artery stenosis rate determination method described in each embodiment of the present invention.

[0071] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A system for determining coronary artery stenosis rate, characterized in that: The invention comprises a processor configured to perform the following method: Performing rough segmentation on the coronary artery image to obtain a rough segmentation result of the coronary artery; Performing fine segmentation on the coarse segmentation result based on the trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and filling the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result; Inputting the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and filling the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result; Determine the coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result; Wherein, determining the coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result includes: performing coronary artery straightening on the coronary artery image, the coronary artery segmentation result and the plaque segmentation result; Determine the position range of each plaque in the coronary artery segmentation result after the coronary artery is straightened, and calculate the stenosis rate of each plaque; The maximum stenosis rate is used as the patient's coronary artery stenosis rate, and the plaque position range corresponding to the coronary artery stenosis rate is mapped to the coronary artery image after the coronary artery is straightened; According to the mapping relationship between the coronary artery image after the coronary artery is straightened and the coronary artery image, the coronary artery stenosis rate and the plaque position range corresponding to the coronary artery stenosis rate are displayed on the coronary artery image; wherein the coronary artery image, the coronary artery segmentation result and the plaque segmentation result have the same size.

2. The system according to claim 1, characterized in that The step of roughly segmenting the coronary artery image to obtain a coarse segmentation result of the coronary artery includes: Downsampling the coronary artery image to a preset coarse segmentation resolution to obtain a first image, and inputting the first image into a trained coarse segmentation model to obtain a coarse segmentation result of the coronary artery; The coarse segmentation result is upsampled to an original resolution to update the coarse segmentation result, wherein the original resolution is the same as a resolution of the coronary artery image.

3. The system according to claim 2, characterized in that After upsampling the coarse segmentation result to the original resolution to update the coarse segmentation result, the method further comprises: The updated coarse segmentation result is expanded to update the coarse segmentation result again.

4. The system according to claim 2, characterized in that The fine segmentation of the coarse segmentation result based on the trained fine segmentation model to obtain the fine segmentation result of the coronary artery includes: The updated coordinates of each pixel in the rough segmentation result and the pixel value of each pixel coordinate in the coronary artery image are input into the trained fine segmentation model to obtain the fine segmentation result of the coronary artery.

5. The system according to claim 1, characterized in that Inputting the coronary artery image into a trained plaque detection model to obtain a plaque prediction result includes: Deducting a coronary artery regional image from the coronary artery image using a coronary artery circumscribed frame corresponding to the fine segmentation result; The coronary artery region image is input into a trained plaque detection model to obtain a plaque prediction result.

6. The system according to claim 5, characterized in that The step of inputting the coronary artery region image into a trained plaque detection model to obtain a plaque prediction result includes: Downsampling the coronary artery region image to a preset plaque segmentation resolution to update the coronary artery region image; Inputting the updated coronary artery region image into the trained plaque detection model to obtain a plaque prediction result; The plaque prediction result is upsampled to an original resolution to update the plaque prediction result, wherein the original resolution is the same as a resolution of the coronary artery image.

7. The system according to claim 1, characterized in that The fine segmentation model is a sparse model.

8. The system according to claim 1, characterized in that Before determining the coronary artery stenosis rate according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result, the method further includes: The blob segmentation result is filtered using the fine segmentation result to update the blob segmentation result.

9. The system according to claim 1, characterized in that The determining of the position range of each plaque in the coronary artery segmentation result after the coronary artery is straightened, and calculating the stenosis rate of each plaque, includes: Based on a preset interval, obtaining the coronary artery lumen diameter in the coronary artery segmentation result after the coronary artery is straightened to obtain a first diameter set; Determine the position range of each plaque in the plaque segmentation result after the coronary artery is straightened in the coronary artery segmentation result after the coronary artery is straightened; Based on the preset interval and the position range of each plaque in the coronary artery segmentation result after the coronary artery is straightened, the coronary artery lumen diameter in the position range of each plaque is obtained to obtain a second diameter set; Determine the lumen diameter at the maximum stenosis of each plaque from the second diameter set to obtain a first lumen diameter subset; determine the proximal normal lumen diameter corresponding to each plaque from the first diameter set to obtain a second lumen diameter subset; The stenosis rate of each plaque is determined according to the first lumen diameter subset and the second lumen diameter subset based on the area method or the diameter method.

10. The system according to claim 1, characterized in that The displaying of the coronary artery stenosis rate and the plaque location range corresponding to the coronary artery stenosis rate on the coronary artery image includes: Using the maximum stenosis rate determined based on the area method as the first coronary artery stenosis rate and / or using the maximum stenosis rate determined based on the diameter method as the second coronary artery stenosis rate; A first coronary artery stenosis rate and a location range of a plaque corresponding to the first coronary artery stenosis rate are displayed on the coronary artery image, and / or a second coronary artery stenosis rate and a location range of a plaque corresponding to the second coronary artery stenosis rate are displayed on the coronary artery image.

11. The system according to claim 1, characterized in that The step of straightening the coronary artery on the coronary artery image, the coronary artery segmentation result and the plaque segmentation result comprises: Determine the coronary artery centerline according to the coronary artery segmentation result; The coronary artery image, the coronary artery segmentation result and the plaque segmentation result are straightened according to the coronary artery centerline.

12. The system according to claim 9, characterized in that Before determining the stenosis rate of each plaque according to the first diameter set and the second diameter set based on the area method and / or the diameter method, the method further includes: The first set of diameters is filtered to update the first set of diameters.

13. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions, when executed by a computer processor, are used to perform the following steps: Performing rough segmentation on the coronary artery image to obtain a rough segmentation result of the coronary artery; Performing fine segmentation on the coarse segmentation result based on the trained fine segmentation model to obtain a fine segmentation result of the coronary artery, and filling the fine segmentation result back into the coronary artery image to obtain a coronary artery segmentation result; Inputting the coronary artery image into a trained plaque detection model to obtain a plaque prediction result, and filling the plaque prediction result back into the coronary artery image to obtain a plaque segmentation result; The coronary artery stenosis rate is determined according to the coronary artery image, the coronary artery segmentation result and the plaque segmentation result.

Images