Standardized cerebrovascular feature extraction method, equipment and medium

Abstract

The invention discloses a standardized cerebrovascular feature extraction method and device and a medium, and relates to the field of cerebrovascular feature extraction, and the method comprises the steps: calculating the blood vessel length and the blood vessel volume based on a target cerebrovascular three-dimensional segmentation image, and determining a blood vessel center line based on Voronoi covariance measurement; according to the blood vessel center line, the blood vessel diameter, the blood vessel curvature radius, the blood vessel bending degree, the blood vessel bifurcation angle and the blood vessel bifurcation density are calculated; performing nonlinear registration on the target cerebrovascular three-dimensional segmented image and a preset standard brain template to obtain a plurality of target standard cerebrovascular partitions and individual feature results in each target standard cerebrovascular partition; and performing group level analysis on the target cerebrovascular three-dimensional segmented image based on a voxel morphological measurement method to obtain a group feature result. According to the invention, multi-level standardized, accurate and stable cerebrovascular feature extraction can be realized.

Description

Technical Field

[0001] The present application relates to the field of cerebrovascular feature extraction, and in particular to a standardized cerebrovascular feature extraction method, device and medium. Background Art

[0002] In the prior art, the signal-to-noise ratios of different cerebral vascular MRI images are different, and the accuracy of different cerebral vascular segmentation methods varies greatly, resulting in unstable and inconsistent feature extraction results. For these problems, there is currently no standardized and effective cerebral vascular feature extraction process. Summary of the invention

[0003] The purpose of this application is to provide a standardized cerebrovascular feature extraction method, device and medium, which can achieve standardized, accurate and stable cerebrovascular feature extraction at individual and group levels.

[0004] To achieve the above objectives, this application provides the following solutions: In a first aspect, the present application provides a standardized cerebral vascular feature extraction method, comprising: Acquire a three-dimensional segmentation image of the target cerebral blood vessels; Based on the target cerebral blood vessel three-dimensional segmentation image, calculating the blood vessel length and blood vessel volume, and determining the blood vessel centerline based on the Voronoi covariance metric; Based on the target cerebral blood vessel three-dimensional segmentation image, the blood vessel diameter, blood vessel curvature radius, blood vessel tortuosity, blood vessel bifurcation angle and blood vessel bifurcation density are calculated according to the blood vessel centerline as individual feature results; Nonlinearly registering the target cerebral blood vessel three-dimensional segmentation image with a preset standard brain template to obtain a plurality of target standard cerebral blood vessel partitions and individual feature results within each target standard cerebral blood vessel partition; the preset standard brain template includes a plurality of preset standard brain partitions; Based on the voxel-based morphological measurement method, the target cerebral blood vessel three-dimensional segmentation image is subjected to group level analysis to obtain group characteristic results.

[0005] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a standardized cerebral vascular feature extraction method.

[0006] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a standardized cerebral vascular feature extraction method.

[0007] According to the specific embodiments provided by the present application, the present application has the following technical effects: the present application provides a standardized cerebrovascular feature extraction method, device and medium, wherein the vascular centerline is determined based on the Voronoi covariance metric, which can assist in the three-dimensional reconstruction of cerebrovascular vessels, thereby reducing the cascade amplification of feature estimation errors caused by small errors in the vascular segmentation process, ensuring the stability and accuracy of vascular feature data, and being able to adapt to the three-dimensional segmentation images of cerebrovascular vessels generated by different segmentation methods. Based on the vascular centerline extracted above, the vascular diameter, vascular curvature radius, vascular bending degree, vascular bifurcation angle and vascular bifurcation density in the target standard cerebrovascular partition, as well as vascular length and vascular volume are calculated as individual feature results. Thus, the present application realizes the mathematical calculation of cerebrovascular basic features, topological features and geometric features on the three-dimensional plane, realizes the automatic extraction of multi-level and multi-dimensional features, and realizes the calculation of standardized personal cerebrovascular image features. Then, the target cerebrovascular three-dimensional segmentation image is nonlinearly registered with the preset standard brain template, and multiple target standard cerebrovascular partitions and individual feature results in each target standard cerebrovascular partition can be obtained. Finally, the voxel-based morphological measurement method performs group-level analysis on the three-dimensional segmentation images of the target cerebral vessels, which has strong representational interpretability. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0009] Figure 1 This is a diagram of the application environment of the standardized cerebral vascular feature extraction method in one embodiment of the present application.

[0010] Figure 2 A schematic diagram of the process of a standardized cerebral vascular feature extraction method provided in one embodiment of the present application.

[0011] Figure 3 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0012] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0013] In order to make the purpose, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0014] The standardized cerebral vascular feature extraction method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the three-dimensional segmentation image of the target cerebral blood vessels to the server 104. After receiving it, the server 104 calculates the blood vessel diameter, blood vessel length, blood vessel curvature radius, blood vessel bending degree, blood vessel bifurcation angle and blood vessel bifurcation density as individual feature results; the three-dimensional segmentation image of the target cerebral blood vessels is analyzed at the group level based on the voxel morphological measurement method to obtain the group feature results. The server 104 can feed back the individual feature results and the group feature results to the terminal 102. In addition, in some embodiments, the standardized cerebral vascular feature extraction method can also be implemented separately by the server 104 or the terminal 102.

[0015] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, and tablet computers. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0016] In an exemplary embodiment, Figure 2 As shown, a standardized cerebrovascular feature extraction method is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used for explanation, and the steps include the following steps 201 to 205.

[0017] Step 201, obtaining a three-dimensional segmented image of a target cerebral blood vessel.

[0018] Step 202, based on the target cerebral blood vessel three-dimensional segmentation image, calculate the blood vessel length and blood vessel volume, and determine the blood vessel centerline based on the Voronoi covariance metric. The calculation process of the blood vessel length and the blood vessel volume includes the following steps (31)-(34).

[0019] (31) Obtaining the individual brain tissue volume corresponding to the target cerebral blood vessel three-dimensional segmentation image by image segmentation; specifically, using the BET (Brain Extraction Tool) tool in the FSL (FMRIB Software Library) toolkit to extract the individual brain tissue volume.

[0020] (32) Counting the number of voxels marked as cerebral blood vessels in the target cerebral blood vessel three-dimensional segmentation image, and calculating the absolute length and absolute volume of the cerebral blood vessels in combination with the voxel resolution. Specifically, the calculation process of the absolute volume of the cerebral blood vessels is as follows: multiplying the number of voxels marked as cerebral blood vessels in the target cerebral blood vessel three-dimensional segmentation image by the voxel volume to obtain the absolute volume of the cerebral blood vessels.

[0021] (33) Considering that the size of individual brain tissue can significantly affect the length of cerebrovascular structure and cerebrovascular volume, the relative length of cerebrovascular is calculated based on the absolute length of cerebrovascular and the volume of individual brain tissue to serve as the vascular length. The specific calculation process is: the absolute length of cerebrovascular is divided by the volume of individual brain tissue to obtain the vascular length.

[0022] (34) Based on the absolute volume of the cerebral blood vessels and the volume of the individual brain tissue, the relative volume of the cerebral blood vessels is calculated as the vascular volume, thereby reflecting the average distribution density of the cerebral blood vessel tissue in the whole brain tissue. The specific calculation process is: the absolute volume of the cerebral blood vessels is divided by the volume of the individual brain tissue to obtain the vascular volume.

[0023] In a specific practical application, the Python-based public software library GDAL and the "create_centerlines" function in centerline are used to stably extract vascular centerlines from the target cerebral vascular 3D segmentation image based on Voronoi Covariance Measure (VCM), thereby reducing the center point identification drift and voxel discontinuity problems at the cerebral vascular bifurcation.

[0024] Step 203, based on the target cerebral blood vessel three-dimensional segmentation image, the blood vessel diameter, blood vessel curvature radius, blood vessel tortuosity, blood vessel bifurcation angle and blood vessel bifurcation density are calculated according to the blood vessel centerline as individual feature results.

[0025] In a specific application, calculating the blood vessel diameter based on the blood vessel centerline includes the following steps (41) to (43).

[0026] (41) Based on the vascular centerline, the λ-maximal segment tangent method (λ-MST) is used to calculate the tangent vector corresponding to each voxel on the vascular centerline. The calculation formula for the tangent vector corresponding to each voxel on the vascular centerline is: .

[0027] in, t ( x ) represents the voxel on the centerline of the vessel x The corresponding tangent vector; P ( x ) indicates that the voxel x The maximum set of line segments, M i is a line segment set P ( x ) represents a line segment; λ() is a preset conversion function, which can be ; t i Represents a line segment M i The tangent vector of e i ( x ) represents a line segment M i Relative voxel x The centrifugal value is calculated as follows: .

[0028] Among them, the molecule Represents the voxel x distance line segment M i Proximal point m i The distance L i Represents line segment M i distance.

[0029] (42) Based on the tangent vector corresponding to each voxel, the normal plane corresponding to each voxel on the centerline of the blood vessel is determined by weighted averaging of the normal planes of neighboring voxels; thus, a robust estimation of the normal plane of each voxel on the centerline of the blood vessel is achieved by weighted averaging of the normal planes of neighboring voxels, that is, a reliable and reasonable result is generated after eliminating noise, outliers or uncertain data. Robust estimation can reduce the impact of bad data (such as noise, outliers, bifurcation complexity, etc.) on the final estimated value, making the result more accurate, smooth, and more consistent with the real data structure.

[0030] (43) For each voxel on the centerline of the blood vessel, a maximum circle is fitted in the normal plane corresponding to the voxel to calculate the corresponding blood vessel diameter.

[0031] Based on the above-obtained vascular centerline and the vascular diameter corresponding to each voxel on the vascular centerline, the three-dimensional reconstruction of the tubular tree-shaped vascular structure can be achieved. Through the above-mentioned vascular diameter calculation steps, it is possible to effectively prevent the local small errors that may be introduced by the structure segmentation algorithm from being cascaded and amplified in the vascular feature estimation, thereby ensuring the robustness of the three-dimensional reconstruction.

[0032] In a specific application, the blood vessel curvature radius and the degree of blood vessel bending are calculated based on the blood vessel centerline, including the following steps (44)-(47).

[0033] (44) Based on the tangent vectors corresponding to each voxel on the centerline of the blood vessel, the angle between the tangent vectors corresponding to adjacent voxels is calculated as the tangent vector flip angle; the tangent vector flip angle represents the direction change between two consecutive voxels; the calculation formula of the tangent vector flip angle is: .

[0034] in, is the vector dot product, is the arccosine function, which is used to calculate the angle between two tangent vectors; is the tangent vector at the ith voxel, is the tangent vector at the i+1th voxel, is the flip angle of the tangent vector at the i-th voxel.

[0035] (45) Calculate the radius of curvature of the blood vessel based on the tangent vector flip angle.

[0036] (46) Based on the centerline of the blood vessel, a Frenet frame is established for each voxel. The Frenet frame is a tool for describing the geometric characteristics of a space curve, which defines a set of orthogonal vectors related to the local characteristics of the curve. It can provide geometric information such as the tangent direction, curvature, and normal vector of the curve at each point. is an orthogonal coordinate system that fully describes the local geometric properties of the curve at each point. The Frenet frame consists of the following three orthogonal vectors: Tangent Vector represents the local direction of the curve (corresponding to the tangent vector in this article), and the principal normal vector N(s) represents the bending direction of the curve, which is defined as: .

[0037] in, is the derivative of the tangent vector, which represents the rate of change of the tangent vector.

[0038] The binormal vector B(s) represents the normal vector of the plane where the curve lies and is defined as: 。 Represents a vector cross product.

[0039] (47) The degree of vascular tortuosity is determined based on the Frenet frame.

[0040] In a specific application, the blood vessel bifurcation angle and blood vessel bifurcation density are calculated based on the blood vessel centerline, including the following steps (48)-(412).

[0041] (48) A tree structure is used to characterize the neighborhood relationship of each voxel on the center line of the blood vessel to construct a three-dimensional tree structure; specifically, a tree structure is used to characterize the 26 neighborhoods of each voxel on the center line of the blood vessel, the adjacency relationship between nodes and edges is defined, and a three-dimensional tree structure is constructed.

[0042] (49) Traverse the adjacency list in the three-dimensional tree structure and count the degree of each voxel.

[0043] (410) If the degree of the voxel is within a preset range, the voxel is marked as a bifurcation point; for example, when the degree is greater than 2, it is a bifurcation point, thereby realizing automatic identification of bifurcation points in the vascular tree structure, and then counting the number of bifurcations and their spatial distribution.

[0044] (411) Calculate the angle between the tangent vectors of the adjacent voxels of the bifurcation point to serve as the corresponding blood vessel bifurcation angle.

[0045] (412) Counting the number of bifurcation points in the target cerebral blood vessel three-dimensional segmentation image, and combining the individual brain tissue volume corresponding to the target cerebral blood vessel three-dimensional segmentation image, calculating the vascular bifurcation density based on the ratio of the number of bifurcation points to the individual brain tissue volume, thereby obtaining the distribution characteristics of the vascular bifurcation statistics in the brain region space. Specifically, the number of bifurcation points is divided by the volume of the target standard cerebral blood vessel partition to obtain the vascular bifurcation density.

[0046] Step 204, nonlinearly aligning the target cerebral blood vessel three-dimensional segmentation image with a preset standard brain template to obtain multiple target standard cerebral blood vessel partitions and individual feature results within each target standard cerebral blood vessel partition; the preset standard brain template includes multiple preset standard brain partitions.

[0047] Specifically, a two-step registration method including linear preliminary registration and nonlinear fine registration is used to nonlinearly register the MNI standard brain template (i.e., the preset standard brain template) with the cerebrovascular magnetic resonance image (i.e., the target cerebrovascular three-dimensional segmentation image) to convert the Harvard-Oxford brain partition map in the MNI standard brain template to the space of the magnetic resonance angiography image at the individual level, thereby dividing the image into multiple partitions, i.e., the target standard cerebrovascular partitions. The individual characteristic results such as vascular length and vascular volume in each target standard cerebrovascular partition can reflect the heterogeneity of the spatial distribution of basic vascular characteristics (including vascular length and vascular volume) in each brain region.

[0048] Step 205, based on voxel based morphometry (VBM), a group level analysis is performed on the target cerebral vascular three-dimensional segmentation image to obtain a group feature result. In a specific application example, step 205 includes the following steps (51)-(53).

[0049] (51) Obtain a cerebrovascular distribution density map at the group level of a healthy population; specifically, the cerebrovascular segmentation images of the healthy population are registered through multiple nonlinear registration iterations to generate a cerebrovascular distribution density map at the group level of a healthy population.

[0050] (52) Nonlinearly aligning the target cerebral vascular three-dimensional segmentation image with the cerebral vascular distribution density map at the level of the healthy population group to obtain a transformation Jacobian matrix.

[0051] (53) Based on the transformation Jacobian matrix, the three-dimensional segmented image of the target cerebral blood vessels after nonlinear registration is normalized to obtain the distribution density of the target cerebral blood vessels in the standard atlas space, and mark it as a group feature result. The distribution density can quantitatively evaluate the difference between the local or global level of the individual vascular distribution network density and the group.

[0052] In summary, this application extracts basic cerebrovascular signs (including vascular length, vascular diameter, and vascular volume), cerebrovascular geometric features (including vascular curvature radius and vascular tortuosity), and cerebrovascular topological features (including vascular bifurcation angle and vascular bifurcation density), and realizes accurate and stable extraction of multi-level standardized cerebrovascular features on cerebrovascular three-dimensional segmentation data, effectively preventing local errors that may be introduced by the structural segmentation algorithm from interfering with vascular feature estimation, that is, this application can effectively suppress noise in segmented images through three-dimensional reconstruction of cerebrovascular segmentation images, and greatly improve the accuracy and consistency of three-dimensional vascular segmentation. This application also adopts VBM analysis and uses iterative registration methods to construct a standard cerebrovascular distribution map of healthy people, transforms and registers individual-level cerebrovascular segmentation images to the standard cerebrovascular distribution space, and realizes comparative analysis of cerebrovascular distribution density at the group level, and has strong representation interpretability.

[0053] In another exemplary application example, in order to systematically solve the problem of feature extraction of 3D segmentation structures of cerebral blood vessels commonly used in clinical and scientific research work, the present application constructs a complete and automatic feature extraction process, which can easily extract a series of objective features with clinical and scientific research significance, such as the diameter, length, curvature, bifurcation density and network characteristics of blood vessels from 3D segmentation images of cerebral blood vessels, quantitatively describe complex structures such as cerebral blood vessels, and propose a standard feature group of cerebral blood vessel structures including the above-mentioned features to assist in related clinical scientific research work.

[0054] This application uses standard spatial registration and VBM methods to establish cerebrovascular characterization statistics and conduct group-level analysis. The former eliminates the interference of local outliers on downstream analysis by weighted average statistics of local brain regions; the latter establishes a standard cerebrovascular template and realizes the conversion from individual to group level by transforming the Jacobian matrix, thereby reducing individual-level noise and achieving robust estimation of cerebrovascular density and group comparative analysis.

[0055] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 3As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a standardized cerebrovascular feature extraction method is implemented.

[0056] Those skilled in the art will understand that Figure 3 The 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 those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0057] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0058] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0059] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0060] Those skilled in the art can understand that the above-mentioned embodiment method is executed by the computer program instruction related hardware, 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 process of the embodiment of each method as described above. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0061] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0062] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above 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.

[0063] The principles and implementation methods of the present application are described herein through specific examples. The description of the above embodiments is only used to help understand the method and core ideas of the present application. At the same time, for those skilled in the art, according to the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A standardized cerebral vascular feature extraction method, characterized in that: The standardized cerebrovascular feature extraction method comprises: Acquire a three-dimensional segmentation image of the target cerebral blood vessels; Based on the target cerebral blood vessel three-dimensional segmentation image, calculating the blood vessel length and blood vessel volume, and determining the blood vessel centerline based on the Voronoi covariance metric; Based on the target cerebral blood vessel three-dimensional segmentation image, the blood vessel diameter, blood vessel curvature radius, blood vessel tortuosity, blood vessel bifurcation angle and blood vessel bifurcation density are calculated according to the blood vessel centerline as individual feature results; Nonlinearly registering the target cerebral blood vessel three-dimensional segmentation image with a preset standard brain template to obtain a plurality of target standard cerebral blood vessel partitions and individual feature results within each target standard cerebral blood vessel partition; the preset standard brain template includes a plurality of preset standard brain partitions; Based on the voxel-based morphological measurement method, the target cerebral blood vessel three-dimensional segmentation image is subjected to group level analysis to obtain group characteristic results.

2. The standardized cerebrovascular feature extraction method according to claim 1, characterized in that: The calculation process of the blood vessel length and the blood vessel volume includes: Acquiring the individual brain tissue volume corresponding to the target cerebral blood vessel three-dimensional segmentation image by image segmentation; Counting the number of voxels marked as cerebral blood vessels in the target 3D segmented image of cerebral blood vessels to calculate the absolute length and absolute volume of the cerebral blood vessels; Calculating the relative length of the cerebral blood vessel as the blood vessel length based on the absolute length of the cerebral blood vessel and the brain tissue volume of the individual; Based on the absolute volume of the cerebral blood vessels and the volume of the individual brain tissue, the relative volume of the cerebral blood vessels is calculated as the blood vessel volume.

3. The standardized cerebrovascular feature extraction method according to claim 1, characterized in that: Calculating the blood vessel diameter according to the blood vessel centerline includes: Based on the blood vessel centerline, a λ-maximum segment tangent method is used to calculate the tangent vector corresponding to each voxel on the blood vessel centerline; Based on the tangent vector corresponding to each voxel, a normal plane corresponding to each voxel on the center line of the blood vessel is determined by weighted averaging of normal planes of adjacent voxels; For each voxel on the centerline of the blood vessel, a maximum circle is fitted in the normal plane corresponding to the voxel to calculate the corresponding blood vessel diameter.

4. The standardized cerebrovascular feature extraction method according to claim 3, characterized in that: The calculation formula of the tangent vector corresponding to each voxel on the center line of the blood vessel is: ； in, t ( x ) represents the voxel on the centerline of the vessel x The corresponding tangent vector; P ( x ) indicates that the voxel x The maximum set of line segments, M i is a line segment set P ( x ) refers to a line segment; λ() is a preset conversion function; t i Represents a line segment M i The tangent vector of e i ( x ) represents a line segment M i Relative voxel x The centrifugal value.

5. The standardized cerebrovascular feature extraction method according to claim 3, characterized in that: Calculating the blood vessel curvature radius and the degree of blood vessel curvature based on the blood vessel centerline includes: According to the tangent vectors corresponding to each voxel on the center line of the blood vessel, the angle between the tangent vectors corresponding to adjacent voxels is calculated as the tangent vector flip angle; the tangent vector flip angle represents the direction change between two consecutive voxels; Calculating the blood vessel curvature radius according to the tangent vector flip angle; According to the centerline of the blood vessel, a Frenet frame of each voxel is established; The degree of blood vessel curvature is determined based on the Frenet frame.

6. The standardized cerebrovascular feature extraction method according to claim 5, characterized in that: The calculation formula of the tangent vector flip angle is: ； in, is the vector dot product, is the arccosine function, is the tangent vector at the ith voxel, is the tangent vector at the i+1th voxel, is the flip angle of the tangent vector at the i-th voxel.

7. The standardized cerebrovascular feature extraction method according to claim 1, characterized in that: Calculating the vascular bifurcation angle and vascular bifurcation density based on the vascular centerline includes: Using a tree structure to characterize the neighborhood relationship of each voxel on the centerline of the blood vessel to construct a three-dimensional tree structure; Traversing the adjacency list in the three-dimensional tree structure, and counting the degree of each voxel; If the degree of the voxel is within a preset range, marking the voxel as a bifurcation point; Calculating the angle of the tangent vectors of the adjacent voxels of the bifurcation point as the corresponding blood vessel bifurcation angle; The number of the bifurcation points in the target cerebral blood vessel three-dimensional segmentation image is counted to calculate the blood vessel bifurcation density.

8. The standardized cerebrovascular feature extraction method according to claim 1, characterized in that: Based on the voxel-based morphological measurement method, the target cerebral blood vessel three-dimensional segmentation image is subjected to group level analysis to obtain group characteristic results, including: Obtain the cerebral vascular distribution density map at the level of healthy people; Nonlinearly registering the target cerebral blood vessel three-dimensional segmentation image with the cerebral blood vessel distribution density map at the level of the healthy population group to obtain a transformation Jacobian matrix; Based on the transformation Jacobian matrix, the three-dimensional segmented image of the target cerebral blood vessel after nonlinear registration is standardized to obtain the distribution density of the target cerebral blood vessel in the standard atlas space, and marked as a group feature result.

9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the standardized cerebrovascular feature extraction method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the standardized cerebrovascular feature extraction method according to any one of claims 1 to 8 is implemented.

Images