Sandwich cavity separation method, device, equipment and medium based on clustering algorithm

Abstract

The present application relates to a method, device, equipment and medium for separating a dissection cavity based on a clustering algorithm, wherein the method for separating a dissection cavity based on a clustering algorithm includes: extracting a vascular dissection endothelial sheet model and a target vascular model based on a CTA image to be processed; converting the vascular dissection endothelial sheet model into a grid model, extracting two continuous and largest-area first and second facets; for each voxel grid on the target vascular model, determining the grid point closest to the first and second facets; calculating the direction vector of the voxel grid and the grid point, and constructing a feature vector with the spatial coordinates of the grid point; performing cluster analysis on the feature vectors corresponding to each voxel grid to obtain a dissection separation result, thereby solving the time-consuming and labor-intensive problem of manual labeling, and having a clear method principle, thereby improving the accuracy of dissection separation.

Description

Technical Field

[0001] The present application relates to the technical field of medical image processing, and in particular to a method, device, equipment and medium for separating interlayer cavities based on a clustering algorithm. Background Art

[0002] The separation of vascular dissection cavities has always been one of the difficult problems in vascular model extraction. Since there may be a large number of flow channels between the dissection cavities, and the development of the endothelial sheet will be affected by CT enhancement, it may not be obvious. For some complex dissection models, the dissection will be distorted, and the difficulty in distinguishing between true and false cavities will be greatly increased. At present, there are two main methods for separating dissection cavities. One is based on manual annotation, and its main disadvantage is that it takes a lot of time and effort; the other is mainly based on deep learning model segmentation methods, and its main disadvantages are as follows: 1. Dependence on a large number of training sets and gold standards; 2. Lack of universality, data trained in a training set is difficult to adapt to images with large differences; 3. Deep learning is a black box, and its application in the medical field requires more caution. Summary of the Invention

[0003] Based on this, it is necessary to provide a method, device, equipment and medium for separating interlayer cavities based on a clustering algorithm to address the above technical problems.

[0004] In a first aspect, an embodiment of the present application provides a method for separating a sandwich cavity based on a clustering algorithm, the method comprising:

[0005] Extracting the vascular dissection endothelium model and the target vessel model based on the CTA image to be processed;

[0006] Converting the vascular dissection endothelial patch model into a mesh model, and extracting two continuous first and second patches with the largest areas;

[0007] For each voxel on the target blood vessel model, determine the grid point closest to the voxel on the first surface patch and the second surface patch; calculate the direction vector between the voxel and the grid point, and construct a feature vector with the spatial coordinates of the grid point;

[0008] Cluster analysis is performed on the eigenvectors corresponding to each voxel grid to obtain the interlayer separation result.

[0009] In one embodiment, converting the vascular dissection endothelial patch model into a mesh model and extracting two continuous first and second patches with the largest areas comprises:

[0010] Obtaining the normal vector of each triangle in the mesh model;

[0011] Dividing the mesh model into a plurality of continuous facets based on the normal vector of each triangular facet;

[0012] Two patches with the largest areas are selected from the plurality of continuous patches to obtain a first patch and a second patch.

[0013] In one embodiment, dividing the mesh model into a plurality of continuous facets based on the normal vectors of the triangular facets comprises:

[0014] Calculating the angle between the normal vectors of each triangle and its adjacent triangles;

[0015] If the normal vector angle is within a threshold range, it is determined that the triangular facet and its adjacent triangular facets are the same continuous facet.

[0016] In one embodiment, for each voxel on the target blood vessel model, determining the grid point closest to the first patch and the second patch; and calculating the direction vector between the voxel and the grid point includes:

[0017] Converting the voxel coordinates of each voxel grid on the target blood vessel model into spatial coordinates;

[0018] Based on the spatial coordinates, calculating the distances between the voxel and all grid points on the first patch and the second patch to determine the closest grid points to the first patch and the second patch;

[0019] Based on the spatial coordinates of the voxel grid and the spatial coordinates of the grid point, a direction vector between the voxel grid and the grid point closest to the voxel grid is calculated.

[0020] In one embodiment, the feature vector V is expressed as follows:

[0021] V=(D, , , ,T);

[0022] ;

[0023] Wherein, D is the direction vector of the voxel and the grid point, x0, y0, z0 are the spatial coordinates of the voxel, x1, y1, z1 are the spatial coordinates of the grid point, when T=1, it means that the grid point is located in the first facet, and when T=2, it means that the grid point is located in the second facet.

[0024] In one embodiment, extracting the endothelial sheet model of the vascular dissection based on the CTA image to be processed includes:

[0025] Extracting an initial vascular model based on the CTA image to be processed;

[0026] Performing intimal sheet filling on the initial blood vessel model based on a morphological closing operation to obtain a first target blood vessel model;

[0027] The first target blood vessel model is subtracted from the initial blood vessel model and the maximum connected domain is obtained to obtain a blood vessel dissection endothelium model.

[0028] In one embodiment, extracting the target blood vessel model based on the CTA image to be processed includes:

[0029] extracting a centerline of the first target blood vessel model;

[0030] Extracting the target point on the center line that is closest to each point of the vascular dissection endothelium model to form a target center line;

[0031] A target blood vessel model is determined based on the target centerline and the first target blood vessel model.

[0032] In a second aspect, an embodiment of the present application further provides a sandwich cavity separation device based on a clustering algorithm, the device comprising:

[0033] A first extraction module is used to extract a vascular dissection endothelium model and a target blood vessel model based on the CTA image to be processed;

[0034] A second extraction module is used to convert the vascular dissection endothelial sheet model into a mesh model, and extract two continuous first and second surface sheets with the largest areas;

[0035] a feature vector construction module, configured to determine, for each voxel on the target blood vessel model, the grid point closest to the first surface patch and the second surface patch; calculate the direction vector between the voxel and the grid point, and construct a feature vector with the spatial coordinates of the grid point;

[0036] The cavity separation module is used to perform cluster analysis on the characteristic vectors corresponding to each voxel grid to obtain the interlayer separation result.

[0037] In a third aspect, an embodiment of the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the method as described in the first aspect above.

[0038] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, wherein the storage medium stores a computer program, wherein when the computer program is executed by a processor, the method described in the first aspect above is implemented.

[0039] The above-mentioned method, device, equipment and medium for separating the dissection cavity based on the clustering algorithm extract the vascular dissection endothelial sheet model and the target vascular model based on the CTA image to be processed; convert the vascular dissection endothelial sheet model into a grid model, and extract two continuous and largest first and second surface patches; for each voxel grid on the target vascular model, determine the grid point closest to the first and second surface patches; calculate the direction vector of the voxel grid and the grid point, and construct a feature vector with the spatial coordinates of the grid point; perform cluster analysis on the feature vectors corresponding to each voxel grid to obtain the dissection separation result, thereby realizing the separation of the dissection cavity based on the traditional segmentation method, solving the time-consuming and labor-intensive manual labeling, and not requiring a large amount of data sets as training sets for deep learning networks. The traditional method has a clear principle and improves the accuracy of cavity separation.

[0040] The details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0042] Figure 1 is a hardware structure block diagram of a terminal device of a sandwich cavity separation method based on a clustering algorithm in an embodiment;

[0043] Figure 2 is a schematic flow chart of a method for separating a sandwich cavity based on a clustering algorithm in one embodiment;

[0044] Figure 3 is a structural block diagram of a sandwich cavity separation device based on a clustering algorithm in one embodiment;

[0045] Figure 4 It is a schematic diagram of the structure of a computer device in an embodiment. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is described and illustrated below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely used to explain this application and are not intended to limit this application. Based on the embodiments provided in this application, all other embodiments obtained by those of ordinary skill in the art without making any creative efforts are within the scope of protection of this application.

[0047] Obviously, the drawings described below are merely examples or embodiments of the present application. Those skilled in the art can, without inventive effort, apply the present application to other similar scenarios based on these drawings. Furthermore, it is also understood that, although the effort involved in such a development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, changes in design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as an insufficiency of the content disclosed in this application.

[0048] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it refer to independent or alternative embodiments that are mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments unless there is a conflict.

[0049] Unless otherwise defined, technical or scientific terms used herein shall have the ordinary meaning as understood by persons of ordinary skill in the art to which this application belongs. The terms "a," "an," "an," "the," and similar expressions used herein do not denote quantitative limitations and may refer to either the singular or the plural. The terms "comprise," "include," "have," and any variations thereof, used herein, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or modules (units) is not limited to the listed steps or units but may also include steps or units not listed, or may include other steps or units inherent to the process, method, product, or apparatus. The terms "connected," "connected," "coupled," and similar expressions used herein are not limited to physical or mechanical connections but may include electrical connections, whether direct or indirect. As used herein, "plurality" means two or more. "And / or" describes an association between associated objects, indicating that three possible relationships exist. For example, "A and / or B" may mean: A exists alone; A and B exist simultaneously; or B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0050] The method embodiment provided in this embodiment can be executed in a terminal, a computer or a similar computing device. For example, running on a terminal, Figure 1FIG. 1 is a block diagram of the hardware structure of the terminal of the sandwich cavity separation method based on the clustering algorithm of this embodiment. Figure 1 As shown, the terminal may include one or more ( Figure 1 Only one is shown) a processor 102 and a memory 104 for storing data, wherein the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA. The above terminal may also include a transmission device 106 and an input and output device 108 for communication functions. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above terminal. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0051] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the interlayer cavity separation method based on the clustering algorithm in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implementing the above-mentioned method. The memory 104 may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely located relative to the processor 102, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0052] The transmission device 106 is used to receive or send data via a network. The network may include a wireless network provided by the terminal's communications provider. In one embodiment, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0053] The present invention provides a method for separating interlayer cavities based on a clustering algorithm, and the method is applied to Figure 1 The terminal in the example is used to illustrate. Figure 2 As shown, the method includes the following steps:

[0054] Step 201 : extracting a vascular dissection endothelium model and a target blood vessel model based on the CTA image to be processed.

[0055] Specifically, an original angiography image is obtained, which is the CTA image to be processed. The embodiment of the present invention extracts the vascular model from the CTA image, specifically by: performing threshold segmentation on the vascular model, marking the vascular model region whose HU value range is within a preset threshold range; using a morphologically based corrosion algorithm to remove small connected regions of bones and blood vessels in the vascular model region, and extracting the connected regions where the blood vessels are located to remove the bone region to obtain the vascular region; using a morphologically based dilation algorithm to compensate the vascular region to obtain an initial vascular model. Then, based on a morphological closing operation, the initial vascular model is filled with an intimal patch to obtain a first target vascular model, and the centerline of the first target vascular model is extracted. The first target vascular model is subtracted from the initial vascular model and the maximum connected domain is taken to obtain an intimal patch model of a vascular dissection.

[0056] For each point on the endothelial sheet model of the vascular dissection, the nearest point on the centerline is found, ultimately yielding a continuous target centerline segment. Using this target centerline segment, only the vascular model near the target centerline is extracted to eliminate interference from distant vascular models in the segmentation process. The target vascular model is then obtained, which serves as the model for the dissection.

[0057] Step 202: Convert the vascular dissection endothelium patch model into a mesh model, and extract two continuous first and second patches with the largest areas.

[0058] Step 203: for each voxel on the target blood vessel model, determine the grid point closest to the first surface patch and the second surface patch; calculate the direction vector between the voxel and the grid point, and construct a feature vector with the spatial coordinates of the grid point.

[0059] Step 204 : performing cluster analysis on the eigenvectors corresponding to the voxel grids to obtain interlayer separation results.

[0060] Specifically, a binary classification clustering algorithm is performed on the feature vectors corresponding to all voxel grids on the target blood vessel model, for example, a k-means algorithm is used for clustering, and the clustering result is the result of the dissection.

[0061] In the above-mentioned method for separating the dissection cavity based on the clustering algorithm, the vascular dissection endothelial sheet model and the target vascular model are extracted based on the CTA image to be processed; the vascular dissection endothelial sheet model is converted into a grid model, and two continuous and largest first and second surface patches are extracted; for each voxel grid on the target vascular model, the grid point closest to the first and second surface patches is determined; the direction vector of the voxel grid and the grid point is calculated, and a feature vector is constructed with the spatial coordinates of the grid point; the feature vectors corresponding to each voxel grid are clustered and analyzed to obtain the dissection separation result. In this way, the separation of the dissection cavity based on the traditional segmentation method is realized, which solves the time-consuming and labor-intensive manual labeling problem while not requiring a large amount of data sets as training sets for deep learning networks. The traditional method has a clear principle and improves the accuracy of dissection separation.

[0062] In one embodiment, converting the vascular dissection endothelial patch model into a mesh model and extracting two continuous first and second patches with the largest areas therefrom comprises the following steps:

[0063] Step 301: Obtain the normal vector of each triangle in the mesh model.

[0064] Step 302: Divide the mesh model into a plurality of continuous facets based on the normal vector of each triangular facet.

[0065] Step 303 : Select two patches with the largest areas from the plurality of continuous patches to obtain a first patch and a second patch.

[0066] In this application, for each triangular facet of the mesh model, its normal vector can be calculated, and then the angle change between the normal vectors of adjacent faces can be analyzed to define whether the two faces are continuous. After analysis, several continuous faces can be obtained (each continuous facet includes multiple triangular faces with small angle changes), and then the two faces with the largest area are taken, namely the first facet and the second facet.

[0067] The first and second surfaces actually represent the two surfaces of the inner membrane sheet, and the definition of the interlayer is actually based on the connection between the inner membrane sheet and the flow cavity. Therefore, it can be judged whether it belongs to a true cavity or a false cavity by the relationship between each point in the flow cavity and the two surfaces of the inner membrane sheet.

[0068] In one embodiment, dividing the mesh model into multiple continuous facets based on the normal vector of each triangular facet includes the following: calculating the angle between the normal vector of each triangular facet and its adjacent triangular facet; if the normal vector angle is within a threshold range, determining that the triangular facet and its adjacent triangular facet are the same continuous facet.

[0069] In this application, if the angle between the normal vectors of a triangle and its adjacent triangles is within a threshold range, for example, less than a certain angle, it means that the angle between the triangle and its adjacent triangles changes little, and the triangle is defined as a continuous patch. Using the above method, the mesh model can be divided into multiple continuous patches, and then the two patches with the largest area are selected.

[0070] In one embodiment, for each voxel on the target blood vessel model, determining the grid point closest to the first patch and the second patch; and calculating the direction vector between the voxel and the grid point comprises the following:

[0071] Step 401: convert the voxel coordinates of each voxel grid on the target blood vessel model into spatial coordinates.

[0072] The voxel grid is the basic discretized representation unit of 3D medical images (such as CTA). In the dissection separation task, the voxel grid is the fundamental unit of calculation and modeling. Its geometric and spatial information directly determines the accuracy of endothelial sheet extraction, feature calculation, and clustering results. The voxel coordinates of each voxel grid are converted into spatial coordinates (x, y, z).

[0073] Step 402 : Based on the spatial coordinates, calculate the distances between the voxel and all grid points on the first patch and the second patch to determine the closest grid points to the first patch and the second patch.

[0074] Specifically, based on the spatial coordinates of the voxel grid and the spatial coordinates of the grid points, the distance between the voxel grid and all grid points on the first patch and the second patch is calculated to determine the grid point closest to the first patch and the second patch. When the closest grid point is on the first patch, it is marked as 1, and when the closest grid point is on the second patch, it is marked as 2.

[0075] Step 403 : Calculate the direction vector between the voxel and the nearest grid point based on the spatial coordinates of the voxel and the spatial coordinates of the grid point.

[0076] In one embodiment, the feature vector V is expressed as follows:

[0077] V=(D, , , ,T);

[0078] ;

[0079] Among them, D is the direction vector of the voxel and the grid point closest to it, x0, y0, z0 are the spatial coordinates of the voxel, x1, y1, z1 are the spatial coordinates of the grid point, when T=1, it means that the grid point is located in the first facet, and when T=2, it means that the grid point is located in the second facet.

[0080] In one embodiment, the process of extracting an intimal patch model of a vascular dissection based on a CTA image to be processed includes: extracting an initial vascular model based on the CTA image to be processed; since the initial vascular model has a disconnected intimal patch, it is difficult to accurately extract the centerline of the vessel; therefore, the intimal patch is filled in using a morphological closing operation; specifically, the intimal patch is filled in the initial vascular model using the morphological closing operation to obtain a first target vascular model; and extracting the centerline of the first target vascular model. The intimal patch model of the vascular dissection is then obtained by subtracting the first target vascular model from the initial vascular model and taking the maximum connected domain.

[0081] In one embodiment, the process of extracting the target vascular model based on the CTA image to be processed includes: extracting the centerline of the first target vascular model; extracting the target point on the centerline that is closest to each point of the vascular dissection endothelium model to form a target centerline; and determining the target vascular model based on the target centerline and the first target vascular model.

[0082] After obtaining the first target vessel model, the present invention extracts the centerline of the first target vessel model. For each point on the endothelial sheet model of the vascular dissection, the nearest point on the centerline is found, ultimately obtaining a continuous target centerline segment. Using this target centerline segment, the method further extracts only the vessel models near the target centerline to eliminate interference from more distant vessel models in the segmentation process, ultimately obtaining the target vessel model. The target vessel model is the model required for dissection.

[0083] It should be noted that the steps shown in the above process or the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0084] In one embodiment, Figure 3 As shown, the embodiment of the present application further provides a sandwich cavity separation device based on a clustering algorithm, the device comprising:

[0085] A first extraction module 10 is configured to extract a vascular dissection endothelium model and a target blood vessel model based on the CTA image to be processed;

[0086] A second extraction module 20 is configured to convert the vascular dissection endothelial sheet model into a mesh model, and extract two continuous first and second surface sheets with the largest areas;

[0087] A feature vector construction module 30 is configured to determine, for each voxel on the target blood vessel model, the grid point closest to the first and second planes; calculate the direction vector between the voxel and the grid point, and construct a feature vector with the spatial coordinates of the grid point;

[0088] The cavity separation module 40 is used to perform cluster analysis on the feature vectors corresponding to each voxel grid to obtain a sandwich separation result.

[0089] In one embodiment, the second extraction module 20 is also used to: obtain the normal vector of each triangular face in the mesh model; divide the mesh model into multiple continuous facets based on the normal vector of each triangular facet; select two facets with the largest area from the multiple continuous facets to obtain a first facet and a second facet.

[0090] In one embodiment, the second extraction module 20 is further configured to calculate the normal vector angle between each triangle and its adjacent triangle; if the normal vector angle is within a threshold range, determine that the triangle and its adjacent triangle are the same continuous facet.

[0091] In one embodiment, the feature vector construction module 30 is further used to: convert the voxel coordinates of each voxel grid on the target blood vessel model into spatial coordinates; based on the spatial coordinates, calculate the distance between the voxel grid and all grid points on the first surface patch and the second surface patch to determine the closest grid point to the first surface patch and the second surface patch; based on the spatial coordinates of the voxel grid and the spatial coordinates of the grid point, calculate the direction vector between the voxel grid and the closest grid point.

[0092] In one embodiment, the feature vector V is expressed as follows:

[0093] V=(D, , , ,T);

[0094] ;

[0095] Wherein, D is the direction vector of the voxel and the grid point, x0, y0, z0 are the spatial coordinates of the voxel, x1, y1, z1 are the spatial coordinates of the grid point, when T=1, it means that the grid point is located in the first facet, and when T=2, it means that the grid point is located in the second facet.

[0096] In one embodiment, the first extraction module 10 is further used to: extract an initial vascular model based on the CTA image to be processed; perform an intimal sheet filling on the initial vascular model based on a morphological closing operation to obtain a first target vascular model; and subtract the first target vascular model from the initial vascular model and take the maximum connected domain to obtain a vascular dissection intimal sheet model.

[0097] In one embodiment, the first extraction module 10 is further used to: extract the center line of the first target blood vessel model; extract the target point on the center line that is closest to each point of the vascular dissection endothelium model to form a target center line; and determine the target blood vessel model based on the target center line and the first target blood vessel model.

[0098] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0099] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 4 The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus.

[0100] The processor of the computer device is used to provide computing and control capabilities. The computer device's memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The computer device's communication interface is used to communicate with an external terminal via wired or wireless communication. Wireless communication can be achieved through Wi-Fi, a mobile cellular network, NFC (near-field communication), or other technologies. When executed by the processor, the computer program implements a clustering algorithm-based interlayer cavity separation method. The display screen of the computer device can be a liquid crystal display or an electronic ink display. The input device of the computer device can be a touch layer covering the display screen, or keys, a trackball, or a touchpad provided on the computer device housing, or an external keyboard, touchpad, or mouse.

[0101] Those skilled in the art will understand that Figure 4The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0102] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned embodiments of the sandwich cavity separation method based on the clustering algorithm are implemented.

[0103] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0104] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0105] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A sandwich cavity separation method based on clustering algorithm, characterized in that: The method comprises: Extracting a vascular dissection endothelial patch model and a target vascular model based on the CTA image to be processed, wherein extracting the vascular dissection endothelial patch model comprises: extracting an initial vascular model based on the CTA image to be processed; performing an intimal patch filling on the initial vascular model based on a morphological closing operation to obtain a first target vascular model; and subtracting the first target vascular model from the initial vascular model and taking the maximum connected domain to obtain the vascular dissection endothelial patch model; Converting the vascular dissection endothelial patch model into a mesh model, and extracting two continuous first and second patches with the largest areas; For each voxel on the target blood vessel model, determine the grid point closest to the voxel on the first surface patch and the second surface patch; calculate the direction vector between the voxel and the grid point, and construct a feature vector with the spatial coordinates of the grid point; Cluster analysis is performed on the eigenvectors corresponding to each voxel grid to obtain the interlayer separation result.

2. The method according to claim 1, characterized in that The step of converting the vascular dissection endothelium patch model into a mesh model and extracting two continuous first and second patches with the largest areas comprises: Obtaining the normal vector of each triangle in the mesh model; Dividing the mesh model into a plurality of continuous facets based on the normal vector of each triangular facet; Two patches with the largest areas are selected from the plurality of continuous patches to obtain a first patch and a second patch.

3. The method according to claim 2, characterized in that The step of dividing the mesh model into a plurality of continuous facets based on the normal vectors of the triangular facets comprises: Calculating the angle between the normal vectors of each triangle and its adjacent triangles; If the normal vector angle is within a threshold range, it is determined that the triangular facet and its adjacent triangular facets are the same continuous facet.

4. The method according to claim 1, wherein The step of determining, for each voxel on the target blood vessel model, the closest grid point to the first surface patch and the second surface patch; and calculating the direction vector between the voxel and the grid point comprises: Converting the voxel coordinates of each voxel grid on the target blood vessel model into spatial coordinates; Based on the spatial coordinates, calculating the distances between the voxel and all grid points on the first patch and the second patch to determine the closest grid points to the first patch and the second patch; Based on the spatial coordinates of the voxel grid and the spatial coordinates of the grid point, a direction vector between the voxel grid and the grid point closest to the voxel grid is calculated.

5. The method according to claim 4, characterized in that The expression of the eigenvector V is as follows: V=(D, , , ,T); ； Wherein, D is the direction vector of the voxel and the grid point, x0, y0, z0 are the spatial coordinates of the voxel, x1, y1, z1 are the spatial coordinates of the grid point, when T=1, it means that the grid point is located in the first facet, and when T=2, it means that the grid point is located in the second facet.

6. The method according to claim 1, characterized in that The extracting of the target blood vessel model based on the CTA image to be processed includes: extracting a centerline of the first target blood vessel model; Extracting the target point on the center line that is closest to each point of the vascular dissection endothelium model to form a target center line; A target blood vessel model is determined based on the target centerline and the first target blood vessel model.

7. A sandwich cavity separation device based on clustering algorithm, characterized in that: The device comprises: a first extraction module, configured to extract a vascular dissection endothelial patch model and a target vascular model based on the CTA image to be processed, wherein extracting the vascular dissection endothelial patch model comprises: extracting an initial vascular model based on the CTA image to be processed; performing an intimal patch filling operation on the initial vascular model based on a morphological closing operation to obtain a first target vascular model; and performing a subtraction between the first target vascular model and the initial vascular model and obtaining a maximum connected domain to obtain the vascular dissection endothelial patch model; A second extraction module is used to convert the vascular dissection endothelial sheet model into a mesh model, and extract two continuous first and second surface sheets with the largest areas; a feature vector construction module, configured to determine, for each voxel on the target blood vessel model, the grid point closest to the first surface patch and the second surface patch; calculate the direction vector between the voxel and the grid point, and construct a feature vector with the spatial coordinates of the grid point; The cavity separation module is used to perform cluster analysis on the characteristic vectors corresponding to each voxel grid to obtain the interlayer separation result.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images