Cardiovascular image analysis method, device, equipment and medium

Abstract

The embodiment of the invention discloses a cardiovascular image analysis method, device and equipment and a medium, and the method comprises the steps: obtaining a to-be-recognized cardiovascular CT image; inputting the to-be-recognized cardiovascular CT image into a pre-trained cardiovascular structure recognition model to obtain a plurality of pieces of target cardiovascular structure position information; and determining the abnormal condition of the coronary artery opening based on the position relationship of the coronary artery opening position in the plurality of pieces of target cardiovascular structure position information relative to the positions of other cardiovascular structures. According to the technical scheme of the embodiment, the problems that the cardiovascular image analysis efficiency is low and the abnormal condition of the coronary artery opening cannot be rapidly analyzed in the prior art are solved, the coronary artery opening point is rapidly detected and the abnormal condition of the coronary artery opening is analyzed, the detection accuracy of the coronary artery opening point can be improved, and meanwhile a plurality of cardiovascular structures can be identified.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of medical image processing, and in particular, to a cardiovascular image analysis method, device, equipment and medium. Background Art

[0002] Computed Tomography angiography (CTA) images are mostly used for the analysis of cardiovascular diseases. Determining the opening positions of the left and right coronary arteries in CTA images can provide important reference information during the analysis of vascular diseases. However, there is currently no fast and accurate CTA image analysis method to identify the opening positions of the left and right coronary arteries. Summary of the Invention

[0003] The embodiments of the present invention provide a cardiovascular image analysis method, device, equipment and medium, which can quickly detect the coronary artery opening points and analyze the abnormal conditions of the coronary artery openings, improve the detection accuracy of the coronary artery opening points, and at the same time identify multiple cardiovascular structures.

[0004] In a first aspect, the embodiments of the present invention provide a cardiovascular image analysis method, which includes:

[0005] Obtain a cardiovascular CT image to be recognized;

[0006] Input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information;

[0007] Based on the positional relationship between the coronary artery opening positions and other cardiovascular structure positions in the multiple target cardiovascular structure position information, determine the abnormal conditions of the coronary artery openings.

[0008] Optionally, based on the positional relationship between the coronary artery opening positions and other cardiovascular structure positions in the multiple target cardiovascular structure position information, determining the abnormal types of the coronary artery openings includes:

[0009] Calculate a first distance between the position of the left coronary sinus and the position of the left coronary artery opening, and a second distance between the position of the right coronary sinus and the position of the right coronary artery opening;

[0010] When the first distance and / or the second distance is greater than a first preset distance threshold, determine that the left coronary sinus and / or the right coronary sinus abnormally opens outside the aortic sinus.

[0011] Optionally, based on the positional relationship between the coronary artery opening positions and other cardiovascular structure positions in the multiple target cardiovascular structure position information, determining the abnormal types of the coronary artery openings further includes:

[0012] Calculate a third distance between the position of the left coronary sinus and the position of the right coronary artery opening, and a fourth distance between the position of the right coronary sinus and the position of the left coronary artery opening;

[0013] When the third distance and / or the fourth distance is less than a second preset distance threshold, it is determined that the left coronary sinus opening and / or the right coronary sinus opening do not match the corresponding left coronary sinus and / or right coronary sinus.

[0014] Optionally, the cardiovascular image analysis method further includes:

[0015] Determine the number of coronary artery opening position information in the target cardiovascular structure position information;

[0016] When the number is greater than two, determine the abnormal situation of the absence of the first cardiovascular structure; and,

[0017] When the number is less than two, determine the abnormal situation of the absence of the second cardiovascular structure.

[0018] Optionally, the training process of the cardiovascular structure recognition model includes:

[0019] Obtain a labeled cardiovascular CT sample image, and generate an attenuated Gaussian sphere centered on the coronary artery opening position based on the coronary artery opening position markers in the cardiovascular CT sample image;

[0020] Input the cardiovascular CT sample image and the attenuated Gaussian sphere into the cardiovascular structure recognition model to be trained for model training to obtain a trained cardiovascular structure recognition model;

[0021] Wherein, the marker information in the cardiovascular CT sample image includes the position information of the coronary artery, left ventricle, left atrium, left coronary sinus, right coronary sinus, left coronary artery opening point and right coronary artery opening point.

[0022] Optionally, the cardiovascular structure recognition model includes a U-shaped network structure, wherein the encoding module in the U-shaped network structure includes a channel attention mechanism and a dimensional attention mechanism.

[0023] Optionally, the cardiovascular image analysis method further includes:

[0024] Mark and display the multiple target cardiovascular structure position information on the cardiovascular CT image to be recognized; and,

[0025] Display preset abnormal prompt information associated with the abnormal situation of the coronary artery opening.

[0026] In a second aspect, an embodiment of the present invention further provides a cardiovascular image analysis device, and the device includes:

[0027] An image acquisition module, configured to acquire a cardiovascular CT image to be recognized;

[0028] A cardiovascular structure recognition module, configured to input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple pieces of target cardiovascular structure position information;

[0029] An abnormality analysis module, configured to determine the abnormality of the coronary artery ostium based on the positional relationship between the coronary artery ostium position in the multiple pieces of target cardiovascular structure position information and other cardiovascular structure positions.

[0030] Optionally, the abnormality analysis module is specifically configured to:

[0031] Calculate a first distance between the position of the left coronary sinus and the position of the left coronary artery ostium, and a second distance between the position of the right coronary sinus and the position of the right coronary artery ostium;

[0032] When the first distance and / or the second distance is greater than a first preset distance threshold, it is determined that the left coronary sinus and / or the right coronary sinus abnormally open outside the aortic sinus.

[0033] Optionally, the abnormality analysis module can also be used to:

[0034] Calculate a third distance between the position of the left coronary sinus and the position of the right coronary artery ostium, and a fourth distance between the position of the right coronary sinus and the position of the left coronary artery ostium;

[0035] When the third distance and / or the fourth distance is less than a second preset distance threshold, it is determined that the left coronary sinus ostium and / or the right coronary sinus ostium do not match the corresponding left coronary sinus and / or right coronary sinus.

[0036] Optionally, the abnormality analysis module can also be used to:

[0037] Determine the number of coronary artery ostium position information in the target cardiovascular structure position information;

[0038] When the number is greater than two, determine the first cardiovascular structure absence abnormality; and,

[0039] When the number is less than two, determine the second cardiovascular structure absence abnormality.

[0040] Optionally, the cardiovascular image analysis device includes a model training module, configured to:

[0041] Obtain a labeled cardiovascular CT sample image, and generate an attenuation Gaussian sphere with the coronary artery ostium position as the center of the sphere based on the coronary artery ostium position label in the cardiovascular CT sample image;

[0042] Input the cardiovascular CT sample image and the attenuation Gaussian sphere into the cardiovascular structure recognition model to be trained for model training to obtain a trained cardiovascular structure recognition model;

[0043] Among them, the marking information in the cardiovascular CT sample image includes the position information of coronary blood vessels, left ventricle, left atrium, left coronary sinus, right coronary sinus, left coronary artery opening point and right coronary artery opening point.

[0044] Optionally, the cardiovascular structure recognition model includes a U-shaped network structure, wherein the encoding module in the U-shaped network structure includes a channel attention mechanism and a dimension attention mechanism.

[0045] Optionally, the cardiovascular image analysis device includes an information display module for:

[0046] Marking and displaying the position information of the multiple target cardiovascular structures on the cardiovascular CT image to be recognized; and,

[0047] Displaying preset abnormal prompt information associated with the abnormal conditions of the coronary artery openings.

[0048] In a third aspect, an embodiment of the present invention further provides a computer device, which includes:

[0049] One or more processors;

[0050] A memory for storing one or more programs;

[0051] When the one or more programs are executed by the one or more processors, the one or more processors implement the cardiovascular image analysis method provided in any embodiment of the present invention.

[0052] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the cardiovascular image analysis method provided in any embodiment of the present invention.

[0053] The technical solution of this embodiment obtains a cardiovascular CT image to be recognized; inputs the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain the position information of multiple target cardiovascular structures; and determines the abnormal conditions of the coronary artery openings based on the positional relationship of the coronary artery opening positions in the position information of the multiple target cardiovascular structures relative to other cardiovascular structure positions. The technical solution of this embodiment solves the problems of low efficiency of cardiovascular image analysis and inability to quickly analyze the abnormal conditions of coronary artery openings in the current clinic, realizes the rapid detection of coronary artery opening points and the analysis of abnormal conditions of coronary artery openings, can improve the detection accuracy of coronary artery opening points, and can identify multiple cardiovascular structures at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flowchart of a cardiovascular image analysis method provided by an embodiment of the present invention;

[0055] Figure 2 The flowchart of a cardiovascular image analysis method provided by an embodiment of the present invention;

[0056] Figure 3 The structural schematic diagram of a cardiovascular image analysis device provided by an embodiment of the present invention;

[0057] Figure 4 The structural schematic diagram of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0058] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. Additionally, it should be noted that for the sake of description, only parts related to the present invention are shown in the drawings, rather than all the structures.

[0059] Figure 1 The flowchart of a cardiovascular image analysis method provided by an embodiment of the present invention. This embodiment is applicable to the scenario of obtaining vascular ultrasound images by signal processing based on intravascular ultrasound signals. This method can be executed by a cardiovascular image analysis device, which can be implemented in software and / or hardware and integrated in a computer device with application development functions.

[0060] As Figure 1 shown, the cardiovascular image analysis method of this embodiment includes the following steps:

[0061] S110. Obtain a cardiovascular CT image to be recognized.

[0062] Among them, the cardiovascular CT image can be a three-dimensional cardiac coronary artery image reconstructed based on angiography data collected by computed tomography.

[0063] The cardiovascular CT image to be recognized can be a three-dimensional cardiac coronary artery image of any imaging scan object.

[0064] Among them, the three-dimensional space coordinate system where the three-dimensional cardiac coronary artery image is located includes the left-right direction of the imaging scan object as the X-axis, and the left half is the positive direction of the X-axis; the chest-to-back direction of the imaging scan object is the positive direction of the Y-axis, and the opposite is the negative direction of the Y-axis; the head-to-foot direction of the imaging scan object is the positive direction of the Z-axis.

[0065] S120. Input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information.

[0066] Among them, the target cardiovascular structure can be one or more structures among coronary blood vessels, left ventricle, left atrium, left coronary sinus, right coronary sinus, left coronary artery opening point, and right coronary artery opening point.

[0067] The cardiovascular structure recognition model is a deep learning network model that has been trained to recognize blood vessels and tissue structure parts in cardiovascular CT images. For example, it can be a Unet model trained with sample images marked with the positions of the above-mentioned multiple target cardiovascular structures.

[0068] S130. Based on the positional relationship of the coronary artery opening position among the multiple target cardiovascular structure position information relative to other cardiovascular structure positions, determine the abnormality of the coronary artery opening.

[0069] Among them, the coronary artery opening refers to the starting part of the coronary artery that supplies blood to the heart. The coronary artery is a group of arteries responsible for transporting oxygen-rich blood to various parts of the heart. The coronary artery openings are usually located at the root of the aorta, that is, the area above the aortic valve. At this position, the coronary artery is divided into the left coronary artery and the right coronary artery. The left coronary artery originates from the left rear of the aortic root and is divided into the left anterior descending branch and the left circumflex branch. The left anterior descending branch supplies blood to the anterior wall, lateral wall, and apex of the left ventricle, and the left circumflex branch supplies blood to the posterior wall, lateral wall, and left atrium of the left ventricle. The right coronary artery originates from the right front of the aortic root and supplies blood to the right ventricle, right atrium, and the inferior wall of the left ventricle. If information such as the positional distribution of the coronary artery openings can be analyzed, it can provide important reference information for the subsequent understanding, analysis, and treatment of heart diseases.

[0070] Specifically, based on the positional relationship of the coronary artery opening position among the multiple target cardiovascular structure position information relative to other cardiovascular structure positions, to determine the abnormal type of the coronary artery opening, it can be to calculate the first distance between the position of the left coronary sinus and the position of the left coronary artery opening, and the second distance between the position of the right coronary sinus and the position of the right coronary artery opening; when the first distance and / or the second distance is greater than the first preset distance threshold, determine that the left coronary sinus and / or the right coronary sinus abnormally open outside the aortic sinus.

[0071] It can also be based on the positional relationship of the coronary artery opening position among the multiple target cardiovascular structure position information relative to other cardiovascular structure positions to determine the abnormal type of the coronary artery opening. It can also be to calculate the third distance between the position of the left coronary sinus and the position of the right coronary artery opening, and the fourth distance between the position of the right coronary sinus and the position of the left coronary artery opening; when the third distance and / or the fourth distance is less than the second preset distance threshold, determine that the left coronary sinus opening and / or the right coronary sinus opening do not match the corresponding left coronary sinus and / or right coronary sinus. It can be understood as an abnormal situation where the coronary artery originates from the contralateral coronary sinus.

[0072] It is understandable that the first distance, the second distance, the third distance, and the fourth distance are distances calculated in a three-dimensional space coordinate system based on multiple target cardiovascular structure position information identified by a cardiovascular structure recognition model.

[0073] In an alternative embodiment, the abnormal condition of the coronary artery opening may also be caused by the absence of cardiovascular structures. When the left circumflex artery and the left anterior descending artery respectively originate from the left coronary sinus and have their own openings, there are three coronary artery openings. When the coronary artery structure is a single coronary artery, that is, the coronary artery lacks the left main trunk or the right main trunk, there is only one coronary artery opening. Therefore, the number of coronary artery opening position information in the target cardiovascular structure position information can be determined; when the number is greater than two, a first cardiovascular structure absence abnormal condition is determined; and when the number is less than two, a second cardiovascular structure absence abnormal condition is determined.

[0074] The technical solution of this embodiment obtains a cardiovascular CT image to be recognized; inputs the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information; and determines the abnormal condition of the coronary artery opening based on the positional relationship of the coronary artery opening position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions. The technical solution of this embodiment solves the problems of low efficiency in cardiovascular image analysis in current clinical practice and the inability to quickly analyze the abnormal condition of the coronary artery opening, realizes the rapid detection of the coronary artery opening point and the analysis of the abnormal condition of the coronary artery opening, can improve the detection accuracy of the coronary artery opening point, and can simultaneously identify multiple cardiovascular structures.

[0075] Figure 2 It is a flowchart of a cardiovascular image analysis method provided by an embodiment of the present invention. This embodiment and the cardiovascular image analysis method in the above embodiment belong to the same inventive concept, and further describes the determination process of the target brightness mapping parameter. This method can be executed by a cardiovascular image analysis device, which can be implemented in a software and / or hardware manner and integrated into a computer device with application development functions.

[0076] As Figure 2 shown, the cardiovascular image analysis method of this embodiment includes the following steps:

[0077] S210. Obtain a marked cardiovascular CT sample image, and generate an attenuation Gaussian sphere with the coronary artery opening position as the center of the sphere based on the coronary artery opening position marking in the cardiovascular CT sample image.

[0078] The marked cardiovascular CT sample image can be a cardiovascular CT image that labels the coronary artery, left ventricle, left atrium, left coronary sinus, right coronary sinus, left coronary artery opening point, and right coronary artery opening point. During the sample image processing, the pixel points in the background area can be labeled as 0; the coronary artery, left ventricle, left atrium, left coronary sinus, and right coronary sinus are labeled as 1, 2, 3, 4, and 5 respectively; the left coronary artery opening point and the right coronary artery opening point are labeled as point coordinates (x1, y1, z1) and (x2, y2, z2).

[0079] Since the coronary artery opening contains more than just one opening point, in order to enable the model to more accurately learn the characteristics of the opening point, an attenuated Gaussian sphere with the coronary artery opening position as the center can be generated in the cardiovascular CT sample image. For example, the relative center position points of the left coronary artery opening and the right coronary artery opening can be selected from the labeled left coronary artery opening point and right coronary artery opening point in the cardiovascular CT sample image. Then, taking the left coronary artery opening point and the right coronary artery opening point as the centers, an attenuated Gaussian sphere with a center of 1 is generated. The Gaussian distribution distributes weights from the inside out, with a weight of 1 at the center of the sphere and a weight of 0 on the surface of the sphere.

[0080] In addition to labeling the left coronary artery opening point and the right coronary artery opening point, the positions of the coronary artery, left ventricle, left atrium, left coronary sinus, and right coronary sinus are also labeled, which can also assist in the cardiovascular structure to identify the left coronary artery opening point and the right coronary artery opening point. Among them, the left coronary artery mainly supplies blood to the anterior myocardium of the left ventricle, while the right coronary artery supplies blood to the posterior myocardium of the left ventricle. Therefore, the coronary artery can be more accurately classified through the left ventricle; the left atrium is connected to the pulmonary veins, and the pulmonary veins will interfere with the segmentation of the coronary artery. Therefore, through the segmentation of the left atrium, the pulmonary veins can be removed from the coronary artery segmentation result.

[0081] After adding the information of the Gaussian sphere, all the sample data sets can also be grouped according to a preset ratio into a training set, a validation set, and a test set.

[0082] S220: Input the cardiovascular CT sample image and the attenuated Gaussian sphere into the cardiovascular structure recognition model to be trained for model training to obtain a trained cardiovascular structure recognition model.

[0083] The cardiovascular structure recognition model includes a U-shaped network, an encoder-decoder-based architecture. In particular, the encoding module (encoder) in the U-shaped network structure in this embodiment includes a channel attention mechanism and a dimensional attention mechanism.

[0084] In this embodiment, since multiple target cardiovascular structures to be recognized are labeled in the cardiovascular CT sample images, it is a multi-target detection task for the model. Adding a channel attention mechanism and a dimensional attention mechanism enables the cardiovascular structure recognition model to better learn the weights of different dimensions of the input data.

[0085] In addition, the advantage of the U-shaped network is that it can effectively handle pixel-level image segmentation tasks, while being able to retain both local and global context information and improve the accuracy and detail preservation ability of the segmentation results through skip connections.

[0086] Input the cardiovascular CT sample images and the corresponding attenuated Gaussian spheres in the training set into the cardiovascular structure recognition model to be trained for model training. When the loss of the cardiovascular structure recognition model converges, the training process can be ended. After testing and verification, the trained cardiovascular structure recognition model can be applied to the corresponding medical image processing scenarios.

[0087] S230. Obtain the cardiovascular CT image to be recognized.

[0088] S240. Input the cardiovascular CT image to be recognized into the pre-trained cardiovascular structure recognition model to obtain the position information of multiple target cardiovascular structures.

[0089] S250. Determine the abnormality of the coronary artery ostium based on the positional relationship of the coronary artery ostium position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions.

[0090] The technical solution of this embodiment is as follows: by obtaining the labeled cardiovascular CT sample images and generating an attenuated Gaussian sphere with the coronary artery ostium position as the center of the sphere based on the labeling of the coronary artery ostium position in the cardiovascular CT sample images; inputting the cardiovascular CT sample images and the attenuated Gaussian spheres into the cardiovascular structure recognition model to be trained for model training to obtain a trained cardiovascular structure recognition model; obtaining the cardiovascular CT image to be recognized; inputting the cardiovascular CT image to be recognized into the pre-trained cardiovascular structure recognition model to obtain the position information of multiple target cardiovascular structures; and determining the abnormality of the coronary artery ostium based on the positional relationship of the coronary artery ostium position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions. The technical solution of this embodiment solves the problems of low efficiency in cardiovascular image analysis in current clinical practice and inability to quickly analyze the abnormality of the coronary artery ostium, realizes the rapid detection of the coronary artery ostium point and the analysis of the abnormality of the coronary artery ostium, can improve the detection accuracy of the coronary artery ostium point, and can also identify multiple cardiovascular structures.

[0091] In an alternative embodiment, after obtaining the position information of multiple target cardiovascular structures and the abnormal conditions of coronary ostia in the cardiovascular CT image to be recognized, the position information of multiple target cardiovascular structures can also be marked and displayed on the cardiovascular CT image to be recognized; and, a preset abnormal prompt information associated with the abnormal conditions of coronary ostia is displayed. That is, visualizing the cardiovascular CT image to be recognized and the recognition result on the interaction interface can provide more intuitive reference information for relevant users.

[0092] In addition, the height of the coronary ostium can be calculated and displayed based on the recognized position information of each structure. Specifically, the annulus plane can be generated by calculating the left coronary sinus and the right coronary sinus in combination with the aortic valve, and the distance from the coronary ostium to the annulus plane can be calculated to obtain the height of the coronary ostium. This information can be used as a reference in transcatheter aortic valve replacement.

[0093] The technical solution in this embodiment can quickly analyze and identify the position information of the left atrium, left ventricle, left coronary sinus, right coronary sinus, and left and right ostium points in the cardiovascular image, which can assist the cardiovascular image analysis work of relevant professionals and improve the processing efficiency of cardiovascular-related problems.

[0094] Figure 3 FIG. 10 is a schematic structural diagram of a cardiovascular image analysis device provided by an embodiment of the present invention. This embodiment is applicable to the scenario of recognizing and analyzing cardiovascular CT images. The device can be implemented in software and / or hardware and integrated into a computer terminal device with application development functions.

[0095] As Figure 3 shown, the cardiovascular image analysis device includes: an image acquisition module 310, a cardiovascular structure recognition module 320, and an abnormality analysis module 330.

[0096] Among them, the image acquisition module 310 is configured to acquire a cardiovascular CT image to be recognized; the cardiovascular structure recognition module 320 is configured to input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain the position information of multiple target cardiovascular structures; the abnormality analysis module 330 is configured to determine the abnormal conditions of the coronary ostia based on the positional relationship of the coronary ostia position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions.

[0097] The technical solution of this embodiment is to obtain a cardiovascular CT image to be recognized; input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information; and determine the abnormal conditions of the coronary artery ostia based on the positional relationship between the coronary artery ostia positions in the multiple target cardiovascular structure position information and other cardiovascular structure positions. The technical solution of this embodiment solves the problems of low efficiency of cardiovascular image analysis in the current clinic and inability to quickly analyze the abnormal conditions of the coronary artery ostia, realizes the rapid detection of the coronary artery ostia points and the analysis of the abnormal conditions of the coronary artery ostia, can improve the detection accuracy of the coronary artery ostia points, and can identify multiple cardiovascular structures at the same time.

[0098] Optionally, the abnormal analysis module 330 is specifically configured to:

[0099] Calculate a first distance between the position of the left coronary sinus and the position of the left coronary artery ostium, and a second distance between the position of the right coronary sinus and the position of the right coronary artery ostium;

[0100] When the first distance and / or the second distance is greater than a first preset distance threshold, it is determined that the left coronary sinus and / or the right coronary sinus abnormally open outside the aortic sinus.

[0101] Optionally, the abnormal analysis module 330 can also be used to:

[0102] Calculate a third distance between the position of the left coronary sinus and the position of the right coronary artery ostium, and a fourth distance between the position of the right coronary sinus and the position of the left coronary artery ostium;

[0103] When the third distance and / or the fourth distance is less than a second preset distance threshold, it is determined that the left coronary sinus ostium and / or the right coronary sinus ostium do not match the corresponding left coronary sinus and / or right coronary sinus.

[0104] Optionally, the abnormal analysis module 330 can also be used to:

[0105] Determine the number of coronary artery ostium position information in the target cardiovascular structure position information;

[0106] When the number is greater than two, determine the abnormal condition of the absence of the first cardiovascular structure; and,

[0107] When the number is less than two, determine the abnormal condition of the absence of the second cardiovascular structure.

[0108] Optionally, the cardiovascular image analysis device includes a model training module for:

[0109] Obtain a labeled cardiovascular CT sample image, and generate an attenuated Gaussian sphere with the coronary artery ostium position as the center of the sphere based on the coronary artery ostium position marking in the cardiovascular CT sample image;

[0110] Input the cardiovascular CT sample image and the attenuated Gaussian sphere into the cardiovascular structure recognition model to be trained for model training, so as to obtain a trained cardiovascular structure recognition model;

[0111] Among them, the marking information in the cardiovascular CT sample image includes the position information of coronary blood vessels, left ventricle, left atrium, left coronary sinus, right coronary sinus, left coronary artery opening point and right coronary artery opening point.

[0112] Optionally, the cardiovascular structure recognition model includes a U-shaped network structure, wherein the encoding module in the U-shaped network structure includes a channel attention mechanism and a dimensional attention mechanism.

[0113] Optionally, the cardiovascular image analysis device includes an information display module for:

[0114] Mark and display the position information of the multiple target cardiovascular structures on the cardiovascular CT image to be recognized; and,

[0115] Display preset abnormal prompt information associated with the abnormal conditions of the coronary artery openings.

[0116] The cardiovascular image analysis device provided by the embodiments of the present invention can execute the cardiovascular image analysis method provided by any embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method.

[0117] Figure 4 It is a schematic structural diagram of a computer device provided by an embodiment of the present invention. Figure 4 It shows a block diagram of an exemplary computer device 12 suitable for implementing the embodiments of the present invention. Figure 4 The shown computer device 12 is only an example, and should not bring any limitation to the functions and usage scope of the embodiments of the present invention. The computer device 12 can be any terminal device with computing power, such as intelligent controllers, servers, mobile phones and other terminal devices.

[0118] Such as Figure 4 shown, the computer device 12 is presented in the form of a general computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0119] Bus 18 represents one or more of several types of bus architectures, including a memory bus or memory controller, a peripheral bus, an Accelerated Graphics Port, a processor bus, or a local bus using any of the several bus architectures. By way of example, and not limitation, these architectures include the Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0120] Computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by computer device 12, including both volatile and nonvolatile media, removable and non-removable media.

[0121] System memory 28 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Computer device 12 can further include other removable / non-removable, volatile / nonvolatile computer system storage media. By way of example only, storage system 34 can be used for reading from and writing to non-removable, nonvolatile magnetic media ( Figure 4 not shown and typically called a "hard disk drive"). Although Figure 4 not shown in the figures, a disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from and writing to a removable, nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) can be provided. In these instances, each drive can be connected to bus 18 by one or more data media interfaces. System memory 28 can include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the present invention.

[0122] A program / utility 40 having a set (at least one) of program modules 42 can be stored, for example, in system memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may include an implementation of a networking environment. The program modules 42 generally carry out the functions and / or methods of the embodiments described herein.

[0123] The computer device 12 can also communicate with one or more external devices 14 (such as keyboards, pointing devices, displays 24, etc.), and can also communicate with one or more devices that enable users to interact with the computer device 12, and / or communicate with any device that enables the computer device 12 to communicate with one or more other computing devices (such as network cards, modems, etc.). Such communication can be carried out through the input / output (I / O) interface 22. Moreover, the computer device 12 can also communicate with one or more networks (such as local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) through the network adapter 20. As shown in the figure, the network adapter 20 communicates with other modules of the computer device 12 through the bus 18. It should be understood that although Figure 4 not shown in the figure, other hardware and / or software modules can be used in combination with the computer device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0124] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the cardiovascular image analysis method provided by the embodiments of the present invention. The method includes:

[0125] Obtain the cardiovascular CT image to be recognized;

[0126] Input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information;

[0127] Based on the positional relationship of the coronary artery ostium position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions, determine the abnormal conditions of the coronary artery ostium.

[0128] The embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the cardiovascular image analysis method provided by any embodiment of the present invention. The method includes:

[0129] Obtain the cardiovascular CT image to be recognized;

[0130] Input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information;

[0131] Based on the positional relationship of the coronary artery ostium position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions, determine the abnormal conditions of the coronary artery ostium.

[0132] The computer storage medium of the embodiments of the present invention may adopt any combination of one or more computer-readable media. The computer-readable media may be computer-readable signal media or computer-readable storage media. The computer-readable storage media may, for example, but not be limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage media include: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this document, the computer-readable storage media may be any tangible medium that contains or stores a program, which can be used by or in combination with an instruction execution system, device, or component.

[0133] The computer-readable signal media may include data signals propagated in a baseband or as part of a carrier wave, which carry computer-readable program codes. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal media may also be any computer-readable medium other than the computer-readable storage media, which can send, propagate, or transmit a program for use by or in combination with an instruction execution system, device, or component.

[0134] The program codes contained on the computer-readable media may be transmitted by any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0135] The computer program codes for performing the operations of the present invention may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program codes may be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0136] Those of ordinary skill in the art should understand that the various modules or steps of the present invention described above can be implemented using a general-purpose computing device. They can be concentrated on a single computing device or distributed across a network composed of multiple computing devices. Optionally, they can be implemented using program code executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0137] Note that the above is only a preferred embodiment of the present invention and the technical principles applied. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein. Various obvious changes, re-adjustments, and substitutions can be made by those skilled in the art without departing from the protection scope 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. Without departing from the concept of the present invention, more other equivalent embodiments can be included, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A cardiovascular image analysis method, characterized in that, Including: Obtain the cardiovascular CT image to be recognized; Input the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information; Based on the positional relationship of the coronary artery ostium position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions, determine the abnormality of the coronary artery ostium.

2. The method according to claim 1, wherein Based on the positional relationship of the coronary artery ostium position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions, determine the abnormal type of the coronary artery ostium, including: Calculate the first distance between the left coronary sinus position and the left coronary artery ostium position, and the second distance between the right coronary sinus position and the right coronary artery ostium position; When the first distance and / or the second distance is greater than the first preset distance threshold, determine that the left coronary sinus and / or the right coronary sinus abnormally open outside the aortic sinus.

3. The method according to claim 2, wherein Based on the positional relationship of the coronary artery ostium position in the multiple target cardiovascular structure position information relative to other cardiovascular structure positions, determining the abnormal type of the coronary artery ostium further includes: Calculate the third distance between the left coronary sinus position and the right coronary artery ostium position, and the fourth distance between the right coronary sinus position and the left coronary artery ostium position; When the third distance and / or the fourth distance is less than the second preset distance threshold, determine that the left coronary sinus ostium and / or the right coronary sinus ostium do not match the corresponding left coronary sinus and / or right coronary sinus.

4. The method according to claim 1, characterized in that, Also including: Determine the number of coronary artery ostium position information in the target cardiovascular structure position information; When the number is greater than two, determine the abnormal situation of the absence of the first cardiovascular structure; And, When the number is less than two, determine the abnormal situation of the absence of the second cardiovascular structure.

5. The method according to any one of claims 1-4, characterized in that, The training process of the cardiovascular structure recognition model includes: Obtain the labeled cardiovascular CT sample image, and generate an attenuated Gaussian sphere with the coronary artery ostium position as the center of the sphere based on the coronary artery ostium position mark in the cardiovascular CT sample image; Input the cardiovascular CT sample image and the attenuated Gaussian sphere into the cardiovascular structure recognition model to be trained for model training to obtain a trained cardiovascular structure recognition model; Wherein, the marking information in the cardiovascular CT sample image includes the position information of coronary arteries, left ventricle, left atrium, left coronary sinus, right coronary sinus, left coronary artery ostium point and right coronary artery ostium point.

6. The method according to claim 5, wherein The cardiovascular structure recognition model includes a U-shaped network structure, wherein the encoding module in the U-shaped network structure includes a channel attention mechanism and a dimensional attention mechanism.

7. The method according to claim 1, characterized in that The method further includes: Mark and display the multiple target cardiovascular structure position information on the cardiovascular CT image to be recognized; and, Display the preset abnormal prompt information associated with the abnormal situation of the coronary artery ostium.

8. A cardiovascular image analysis device, characterized in that, Including: An image acquisition module for obtaining the cardiovascular CT image to be recognized; A cardiovascular structure recognition module for inputting the cardiovascular CT image to be recognized into a pre-trained cardiovascular structure recognition model to obtain multiple target cardiovascular structure position information; An abnormality analysis module, configured to determine an abnormality of a coronary artery opening based on a positional relationship between the coronary artery opening position in the plurality of target cardiovascular structure position information and other cardiovascular structure positions.

9. A computer device, characterized in that, The computer device includes: One or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the cardiovascular image analysis method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the cardiovascular image analysis method according to any one of claims 1-7.

Images