Method and device for extracting center line of blood vessel

Through the Hessian matrix filter and DBSCAN clustering algorithm, combined with bifurcation recognition processing, the problem of confusion between blood vessels and bones is solved, the extraction accuracy of the blood vessel center line is improved, and the blood vessel extraction effect in complex images is achieved.

CN119068044APending Publication Date: 2024-12-03SUZHOU AMIMED MEDICAL TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411100017.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-12
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

Existing vascular extraction methods can easily confuse blood vessels and bones in medical images with complex spatial locations, resulting in failure of vascular segmentation and inability to extract the complete vascular center line.

Method used

The Hessian matrix filter is used to process the local area of ​​the blood vessel, obtain the feature vector of the blood vessel characteristics, and map it to the preset sphere. The DBSCAN clustering algorithm is used to calculate the center point of the local blood vessel, and the center line of the blood vessel is determined by combining the sampling center point. DBSCAN algorithm parameters and bifurcation recognition processing improve the extraction accuracy.

Benefits of technology

It improves the extraction accuracy of the blood vessel center line, and can more accurately extract blood vessels under small, low contrast and fracture, enhancing the ability to capture blood vessel geometric topology.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119068044A_ABST
    Figure CN119068044A_ABST
Patent Text Reader

Abstract

The embodiment of the invention provides a blood vessel center line extraction method and device, and the method comprises the steps: obtaining a blood vessel local region image which is obtained through the sampling of an original image of a blood vessel; processing the blood vessel local region image based on a Hessian matrix filter to obtain a feature vector of blood vessel features; the feature vector is mapped to a preset spherical surface, a DBSCAN clustering algorithm is used to calculate the central point of the local blood vessel, and the center of the preset spherical surface is a sampling central point; and determining a blood vessel center line based on the center point and the sampling center point. The embodiment of the invention can improve the precision of extracting the center line of the blood vessel.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of medical image processing, and in particular to a method and device for extracting a blood vessel centerline. Background Art

[0002] Vascular imaging technologies, such as computed tomography angiography (CTA), magnetic resonance angiography (MRA), and magnetic resonance imaging without contrast agents, provide doctors with important tools for diagnosing vascular-related diseases, including vascular calcification, stenosis, aneurysm, and dura mater lesions. Although these technologies can generate three-dimensional images of blood vessels, these images themselves are not enough for doctors to gain an intuitive understanding. Therefore, it is necessary to further extract the morphological structure of blood vessels from medical images and display the morphology of these blood vessels through three-dimensional visualization technology. Such a display method helps doctors analyze the lesion area more accurately, thereby improving the accuracy and scientificity of the diagnosis. This is crucial for formulating the best treatment strategy and surgical plan, and has a far-reaching impact on medical research.

[0003] However, in actual blood vessel extraction, due to the complex spatial position and the overlap of the enhanced vascular CT values ​​and bone CT values, common blood vessel extraction methods, such as region growing and level set, can easily confuse blood vessels and bones, resulting in blood vessel segmentation failure and inability to extract the complete blood vessel centerline.

[0004] Therefore, there is an urgent need to provide a method for extracting the centerline of a blood vessel, thereby improving the accuracy of extracting the centerline of the blood vessel. Summary of the invention

[0005] In view of this, the method and device for extracting the centerline of a blood vessel provided in the embodiments of the present invention can improve the accuracy of extracting the centerline of a blood vessel.

[0006] To achieve the above objectives, in a first aspect, an embodiment of the present invention provides a method for extracting a blood vessel centerline, comprising:

[0007] Acquire a local area image of the blood vessel, wherein the local area image of the blood vessel is obtained by sampling an original image of the blood vessel;

[0008] The local area image of the blood vessel is processed based on the Hessian matrix filter to obtain the characteristic vector of the blood vessel characteristics;

[0009] Mapping the feature vector to a preset sphere, and using the DBSCAN clustering algorithm to calculate the center point of the local blood vessel, the center of the preset sphere being the sampling center point;

[0010] Based on the center point and the sampling center point, the blood vessel center line is determined.

[0011] Optionally, the using of the DBSCAN clustering algorithm to calculate the center point of the local blood vessel includes:

[0012] Adjusting the parameters of the DBSCAN clustering algorithm to determine a first category of a preset number of local blood vessels, the first category being a category in which a proportion of categories corresponding to blood vessels is higher than a preset value;

[0013] A central point is calculated for all points in the first category as the central point of the local blood vessel.

[0014] Optionally, determining the blood vessel centerline based on the center point and the sampling center point includes:

[0015] The calculation center point and the sampling center point form the travel direction;

[0016] Determine the target sampling center point whose angle difference between the direction vector formed by the sampling center point and the current sampling center point and the traveling direction is less than or equal to a preset angle;

[0017] Determine whether the score corresponding to the current blood vessel local area image is not less than a preset score;

[0018] If so, bifurcation identification is performed to determine the blood vessel centerline.

[0019] Optionally, performing bifurcation identification processing to determine the blood vessel centerline includes:

[0020] Determine a bifurcation point after bifurcation identification processing, wherein the bifurcation point is an intersection point where at least two vascular branches exist in the target sampling center point;

[0021] Based on the center point and bifurcation point, the blood vessel centerline is determined.

[0022] Optionally, the preset angle is 20°; the preset score is a score, and the score is a score not greater than 0.01.

[0023] In a second aspect, an embodiment of the present invention provides a device for extracting a blood vessel centerline, comprising:

[0024] An acquisition module, used for acquiring a local area image of a blood vessel, wherein the local area image of the blood vessel is obtained by sampling an original image of the blood vessel;

[0025] Hessian matrix filter, used to process the image of the local area of ​​the blood vessel to obtain the characteristic vector of the blood vessel characteristics;

[0026] A mapping module, used to map the feature vector to a preset sphere and calculate the center point of the local blood vessel using a DBSCAN clustering algorithm, wherein the center of the preset sphere is the sampling center point;

[0027] The calculation module is used to determine the center line of the blood vessel based on the center point and the sampling center point.

[0028] Optionally, the mapping module is used to calculate the center point of the local blood vessel using the DBSCAN clustering algorithm, including:

[0029] Adjusting the parameters of the DBSCAN clustering algorithm to determine a first category of a preset number of local blood vessels, the first category being a category in which a proportion of categories corresponding to blood vessels is higher than a preset value;

[0030] A central point is calculated for all points in the first category as the central point of the local blood vessel.

[0031] Optionally, the calculation module is used to determine the blood vessel centerline based on the center point and the sampling center point, including:

[0032] The calculation center point and the sampling center point form the travel direction;

[0033] Determine the target sampling center point whose angle difference between the direction vector formed by the sampling center point and the current sampling center point and the traveling direction is less than or equal to a preset angle;

[0034] Determine whether the score corresponding to the current blood vessel local area image is not less than a preset score;

[0035] If so, bifurcation identification processing is performed to determine the centerline of the blood vessel.

[0036] Optionally, the calculation module is used to perform bifurcation identification processing to determine the blood vessel centerline, including:

[0037] Determine a bifurcation point after bifurcation identification processing, wherein the bifurcation point is an intersection point where at least two vascular branches exist in the target sampling center point;

[0038] Based on the center point and bifurcation point, the blood vessel centerline is determined.

[0039] Optionally, the preset angle is 20°; the preset score is a score, and the score is a score not greater than 0.01.

[0040] Based on the above, an embodiment of the present invention provides a method for extracting the centerline of a blood vessel, including: obtaining a local area image of a blood vessel, wherein the local area image of the blood vessel is obtained by sampling the original image of the blood vessel; processing the local area image of the blood vessel based on a Hessian matrix filter to obtain a characteristic vector of the blood vessel characteristics; mapping the characteristic vector to a preset sphere, and using a DBSCAN clustering algorithm to calculate the center point of the local blood vessel, wherein the center of the preset sphere is the sampling center point; and determining the centerline of the blood vessel based on the center point and the sampling center point. Based on the above, an embodiment of the present invention uses the characteristic vector calculated by the Hessian matrix filter to convert the directional characteristics of the blood vessel, and then through the clustering algorithm, the directional information of each iteration can be captured, thereby obtaining the centerline of the blood vessel and the geometric topological structure. Compared with similar methods, the embodiment of the present invention can extract more blood vessel levels, improve the accuracy of extracting the centerline of the blood vessel, and at the same time ensure the extraction effect of blood vessels in small, low-contrast, broken and other conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0042] Figure 1 A schematic diagram of the steps of a method for extracting a blood vessel centerline provided in an embodiment of the present invention.

[0043] Figure 2 A schematic diagram of a feature vector of a blood vessel feature provided by an embodiment of the present invention.

[0044] Figure 3 An optional schematic diagram of rendering a blood vessel provided by an embodiment of the present invention.

[0045] Figure 4 An optional schematic diagram of blood vessel mapping provided by an embodiment of the present invention.

[0046] Figure 5 A schematic diagram of calculating the center point of a local blood vessel provided by an embodiment of the present invention.

[0047] Figure 6 Another optional schematic diagram of blood vessel mapping provided by an embodiment of the present invention.

[0048] Figure 7 This is an extraction comparison diagram of the blood vessel centerline extraction method provided in an embodiment of the present invention.

[0049] Figure 8A block diagram of a device for extracting a blood vessel centerline provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of this application.

[0051] The embodiment of the present invention provides a method for extracting the centerline of a blood vessel, comprising: obtaining a local area image of the blood vessel, wherein the local area image of the blood vessel is obtained by sampling the original image of the blood vessel; processing the local area image of the blood vessel based on a Hessian matrix filter to obtain a characteristic vector of the blood vessel characteristics; mapping the characteristic vector to a preset sphere, and using a DBSCAN clustering algorithm to calculate the center point of the local blood vessel, wherein the center of the preset sphere is the sampling center point; and determining the centerline of the blood vessel based on the center point and the sampling center point. Based on the above, the embodiment of the present invention uses the characteristic vector calculated by the Hessian matrix filter to convert the directional characteristics of the blood vessel, and then through the clustering algorithm, the directional information of each iteration can be captured, thereby obtaining the centerline of the blood vessel and the geometric topological structure. Compared with similar methods, the embodiment of the present invention can extract more blood vessel levels, improve the accuracy of extracting the centerline of the blood vessel, and at the same time ensure the extraction effect of blood vessels in small, low-contrast, broken and other conditions.

[0052] Specifically, the method for extracting the centerline of a blood vessel according to an embodiment of the present invention can be referred to Figure 1 As shown, it specifically includes:

[0053] Step S1, obtaining a local area image of a blood vessel, wherein the local area image of the blood vessel is obtained by sampling an original image of the blood vessel;

[0054] The local area image of the blood vessel is obtained by sampling the original image of the blood vessel, including: extracting an original feature map from the original image of the blood vessel, determining the probability value of each pixel point in the image data belonging to the blood vessel based on the characteristics of each pixel in the original feature map, and forming a local area map of the blood vessel of the image data through sampling.

[0055] Step S2, processing the local area image of the blood vessel based on the Hessian matrix filter to obtain a feature vector of the blood vessel feature;

[0056] The Hessian matrix filter in the embodiment of the present invention can capture local changes in the image, identify features such as corner points and edges in the image, and thus determine the feature vector of the blood vessel feature.

[0057] In an optional embodiment, the intensity value at the coordinate x is recorded as I(x), where I(x) represents the intensity at the coordinate x=[x1,…,x D ] T The intensity value at x1,…,x D Corresponding to the coordinates in each dimension, T is the vector transposition symbol, through x1,…,x D It can locate the specific pixel position in the D-dimensional image, and the value stored at this position is I(x).

[0058] Optionally, when the D-dimensional image is a picture, it is a two-dimensional image, and when the D-dimensional image is a contrast image, it is a three-dimensional image.

[0059] Among them, the intensity at the above coordinate x is I(x) and the Hessian matrix of scale s is a D×D matrix:

[0060]

[0061] Where G(x,s)=(2πs 2 ) -D / 2 exp(-x T x / 2s 2 )Formula 2

[0062] H ij (x,s) means the coordinates x=[x1,…,x D ] T The second-order partial derivative matrix at coordinates x = [x1,…,x D ] T For a Gaussian convolution kernel with a scale of s, the embodiment of the present invention provides a 3×3 Hessian matrix as described in Formula 3:

[0063]

[0064] Among them, * is convolution.

[0065] The calculation process of the above eigenvalues ​​is obtained by the following formula:

[0066] |λE-H|=0Formula 4

[0067] Among them, λ is the eigenvalue. The third-order matrix can obtain three eigenvalues, which are sorted from small to large, namely λ1, λ2, and λ3. Among them, the eigenvector of λ1 corresponds to υ1, the eigenvector of λ2 corresponds to υ2, and the eigenvector of λ3 corresponds to υ3.

[0068] Specifically, the feature vector of the vascular feature can be referred to Figure 2As shown, υ1 represents the axial direction of the vascular tubular structure, υ2 and υ3 represent the cross-sectional directions of the tubular structure, υ2 represents the vertical direction of the cross-sectional direction of the vascular tubular structure, the vertical angle between any two of υ1, υ2, and υ3 is 90 degrees, and υ3 represents the depth direction of the cross-sectional direction of the vascular tubular structure.

[0069] Furthermore, after obtaining the axial direction υ1, a three-dimensional rendering visualization is performed to obtain the following Figure 3 The schematic diagram shown in Figure 1 shows the schematic diagram of the present invention. Figure 3 (a) is the vascular direction feature of a single simple blood vessel local sampling; Figure 3 (b) is the vascular direction feature of local sampling of a single vessel with a bend; Figure 3 (c) is the blood vessel direction feature of the local sampling of the bifurcation.

[0070] Step S3, mapping the feature vector to a preset sphere, and using the DBSCAN clustering algorithm to calculate the center point of the local blood vessel, the center of the preset sphere being the sampling center point;

[0071] Specifically, the eigenvector υ1 calculated by the Hessian matrix of all coordinate positions with a response value of 1.0 in the sampling area can be mapped to a preset spherical surface with the sampling center point as the sphere center, that is, converted into a point cloud such as Figure 4 As shown, Figure 4 (a) is a schematic diagram of the mapping of straight blood vessels. Figure 4 (b) is a schematic diagram of the mapping of blood vessels at the bend.

[0072] In an optional embodiment, the central point of the local blood vessel is calculated using the DBSCAN clustering algorithm, such as Figure 5 As shown, including:

[0073] Step S51, adjusting the parameters of the DBSCAN clustering algorithm to determine a first category of a preset number of local blood vessels, where the first category is a category corresponding to blood vessels with a proportion higher than a preset value;

[0074] Step S52: Calculate a central point for all points in the first category as the central point of the local blood vessel.

[0075] In an optional implementation, the distance threshold in the DBSCAN algorithm parameters is set to 0.5, the minimum number of data points is set to 10, and then DBSCAN clustering is performed; after DBSCAN clustering, the top five categories in terms of category proportion are selected, and then the center points of all points in these five categories are calculated, and then the center points of all points in the largest category are calculated. The vector formed by the center point and the sampling center point is the blood vessel direction for the next iteration.

[0076] Step S4: Determine the blood vessel centerline based on the center point and the sampling center point.

[0077] Optionally, determining the blood vessel centerline based on the center point and the sampling center point includes:

[0078] The calculation center point and the sampling center point form the travel direction;

[0079] Determine the target sampling center point whose angle difference between the direction vector formed by the sampling center point and the current sampling center point and the traveling direction is less than or equal to a preset angle;

[0080] In an optional embodiment, the preset angle is 20°, 30°, etc.

[0081] Determine whether the score corresponding to the current blood vessel local area image is not less than a preset score;

[0082] The preset score is a score, and the score is a score not greater than 0.01.

[0083]

[0084] The score is calculated by counting the number of response values ​​that are equal to 1.0 and the number of response values ​​that are not equal to 1.0, and comparing the number that is equal to 1.0 to the total.

[0085] Res(λ1,λ2,λ3)=1.0, if true, the count increases by one;

[0086]

[0087] λ ρ (s) represents the maximum eigenvalue among the eigenvalues ​​λ3 at the corresponding scale s at all positions in the image.

[0088] τ is an introduced disturbance variable, between 0 and 1, and is usually set to 0.75.

[0089]

[0090] Among them, Res(λ1,λ2,λ3) can be obtained by the above formula, and the score as described above is obtained based on the response value corresponding to Res(λ1,λ2,λ3).

[0091] If so, bifurcation identification is performed to determine the blood vessel centerline.

[0092] Optionally, performing bifurcation identification processing to determine the blood vessel centerline includes:

[0093] Determine a bifurcation point after bifurcation identification processing, wherein the bifurcation point is an intersection point where at least two vascular branches exist in the target sampling center point;

[0094] Based on the center point and bifurcation point, the blood vessel centerline is determined.

[0095] In an embodiment of the present invention, Figure 6 (a) is a schematic diagram of the mapping of blood vessels with a large bifurcation angle. Figure 6 (b) is a schematic diagram of the mapping of blood vessels with a smaller bifurcation angle.

[0096] The present invention further provides experimental results of vascular center extraction under different parameters, such as Figure 7 As shown, Figure 7 (a) is an extraction effect diagram of the blood vessel centerline extraction method provided in this application, Figure 7 (b) is the extraction effect diagram of the existing blood vessel centerline extraction method.

[0097] The embodiment of the present invention also provides a corresponding blood vessel centerline extraction device, see Figure 8 The device for extracting the centerline of a blood vessel shown in the figure, wherein the device for extracting the centerline of a blood vessel may include:

[0098] An acquisition module 81 is used to acquire a local area image of a blood vessel, wherein the local area image of the blood vessel is obtained by sampling an original image of the blood vessel;

[0099] Hessian matrix filter 82, used for processing the local area image of the blood vessel to obtain the characteristic vector of the blood vessel characteristics;

[0100] A mapping module 83 is used to map the feature vector to a preset sphere and calculate the center point of the local blood vessel using a DBSCAN clustering algorithm, wherein the center of the preset sphere is the sampling center point;

[0101] The calculation module 84 is used to determine the blood vessel centerline based on the center point and the sampling center point.

[0102] In an optional embodiment, the mapping module is used to calculate the center point of the local blood vessel using the DBSCAN clustering algorithm, including:

[0103] Adjusting the parameters of the DBSCAN clustering algorithm to determine a first category of a preset number of local blood vessels, the first category being a category in which a proportion of categories corresponding to blood vessels is higher than a preset value;

[0104] A central point is calculated for all points in the first category as the central point of the local blood vessel.

[0105] In an optional embodiment, the calculation module is used to determine the blood vessel centerline based on the center point and the sampling center point, including:

[0106] The calculation center point and the sampling center point form the travel direction;

[0107] Determine the target sampling center point whose angle difference between the direction vector formed by the sampling center point and the current sampling center point and the traveling direction is less than or equal to a preset angle;

[0108] Determine whether the score corresponding to the current blood vessel local area image is not less than a preset score;

[0109] If so, bifurcation identification processing is performed to determine the centerline of the blood vessel.

[0110] In an optional embodiment, the calculation module is used to perform bifurcation identification processing to determine the blood vessel centerline, including:

[0111] Determine a bifurcation point after bifurcation identification processing, wherein the bifurcation point is an intersection point where at least two vascular branches exist in the target sampling center point;

[0112] Based on the center point and bifurcation point, the blood vessel centerline is determined.

[0113] In an optional embodiment, the preset angle is 20°; the preset score is a score, and the score is a score not greater than 0.01.

[0114] The above describes multiple implementation schemes provided by the embodiments of the present invention. The various optional methods introduced in the implementation schemes can be combined and cross-referenced with each other without conflict, thereby extending a variety of possible implementation schemes, which can all be considered as implementation schemes disclosed and open in the embodiments of the present invention.

[0115] Although the embodiments of the present invention are disclosed above, the present application is not limited thereto. Any person skilled in the art may make various changes and modifications without departing from the spirit and scope of the present application. Therefore, the protection scope of the present application shall be subject to the scope defined by the claims.

Claims

1. A method for extracting a blood vessel centerline, characterized in that: include: Acquire a local area image of the blood vessel, wherein the local area image of the blood vessel is obtained by sampling an original image of the blood vessel; The local area image of the blood vessel is processed based on the Hessian matrix filter to obtain the characteristic vector of the blood vessel characteristics; Mapping the feature vector to a preset sphere, and using the DBSCAN clustering algorithm to calculate the center point of the local blood vessel, the center of the preset sphere being the sampling center point; Based on the center point and the sampling center point, the blood vessel center line is determined.

2. The method for extracting the blood vessel centerline according to claim 1, characterized in that: The method of using the DBSCAN clustering algorithm to calculate the center point of the local blood vessel includes: Adjusting the parameters of the DBSCAN clustering algorithm to determine a first category of a preset number of local blood vessels, the first category being a category in which the proportion of categories corresponding to the blood vessels is higher than a preset value; A central point is calculated for all points in the first category as the central point of the local blood vessel.

3. The method for extracting the blood vessel centerline according to claim 1, characterized in that: The step of determining the blood vessel centerline based on the center point and the sampling center point includes: The calculation center point and the sampling center point form the travel direction; Determine the target sampling center point whose angle difference between the direction vector formed by the sampling center point and the current sampling center point and the traveling direction is less than or equal to a preset angle; Determine whether the score corresponding to the current blood vessel local area image is not less than a preset score; If so, bifurcation identification processing is performed to determine the centerline of the blood vessel.

4. The method for extracting the blood vessel centerline according to claim 3, characterized in that: The bifurcation identification process is performed to determine the blood vessel centerline, including: Determine a bifurcation point after bifurcation identification processing, wherein the bifurcation point is an intersection point where at least two vascular branches exist in the target sampling center point; Based on the center point and bifurcation point, the blood vessel centerline is determined.

5. The method for extracting the blood vessel centerline according to claim 4, characterized in that: The preset angle is 20°; the preset score is a score, and the score is a score not greater than 0.

01.

6. A device for extracting a blood vessel centerline, characterized in that: include: An acquisition module, used for acquiring a local area image of a blood vessel, wherein the local area image of the blood vessel is obtained by sampling an original image of the blood vessel; Hessian matrix filter, used to process the image of the local area of ​​the blood vessel to obtain the characteristic vector of the blood vessel characteristics; A mapping module, used to map the feature vector to a preset sphere and calculate the center point of the local blood vessel using a DBSCAN clustering algorithm, wherein the center of the preset sphere is the sampling center point; The calculation module is used to determine the center line of the blood vessel based on the center point and the sampling center point.

7. The device for extracting the blood vessel centerline according to claim 1, characterized in that: The mapping module is used to calculate the center point of the local blood vessel using the DBSCAN clustering algorithm, including: Adjusting the parameters of the DBSCAN clustering algorithm to determine a first category of a preset number of local blood vessels, the first category being a category in which the proportion of categories corresponding to the blood vessels is higher than a preset value; A central point is calculated for all points in the first category as the central point of the local blood vessel.

8. The device for extracting the blood vessel centerline according to claim 6, characterized in that: The calculation module is used to determine the blood vessel centerline based on the center point and the sampling center point, including: The calculation center point and the sampling center point form the travel direction; Determine the target sampling center point whose angle difference between the direction vector formed by the sampling center point and the current sampling center point and the traveling direction is less than or equal to a preset angle; Determine whether the score corresponding to the current blood vessel local area image is not less than a preset score; If so, bifurcation identification processing is performed to determine the centerline of the blood vessel.

9. The device for extracting the blood vessel centerline according to claim 8, characterized in that: The calculation module is used to perform bifurcation identification processing and determine the blood vessel centerline, including: Determine a bifurcation point after bifurcation identification processing, wherein the bifurcation point is an intersection point where at least two vascular branches exist in the target sampling center point; Based on the center point and bifurcation point, the blood vessel centerline is determined.

10. The device for extracting the blood vessel centerline according to claim 9, characterized in that: The preset angle is 20°; the preset score is a score, and the score is a score not greater than 0.01.