CN111986205A

CN111986205A - Vessel tree generation and lesion identification method, device, equipment and readable storage medium

Info

Publication number: CN111986205A
Application number: CN201910425253.6A
Authority: CN
Inventors: 梁红霞; 赵丽俊; 张晓雅; 董爱莲
Original assignee: Individual
Current assignee: Individual
Priority date: 2019-05-21
Filing date: 2019-05-21
Publication date: 2020-11-24
Anticipated expiration: 2039-05-21
Also published as: CN111986205B

Abstract

The invention discloses a method, a device and equipment for generating a blood vessel tree and identifying lesions and a readable storage medium. The method comprises the following steps: acquiring a medical image, the medical image comprising a plurality of voxels; segmenting a plurality of vessel regions from the medical image, the vessel regions comprising: blood vessels and voxels with similar characteristics to blood vessels; establishing an incidence matrix of the plurality of blood vessel regions according to the adjacent relation among the plurality of blood vessel regions; respectively acquiring a blood vessel section and a blood vessel bifurcation point of each blood vessel region; constructing the vessel segments, the vessel bifurcation points and the connection relationship among the vessel segments and the vessel bifurcation points contained in each vessel region into a vessel map of the plurality of vessel regions according to the incidence matrix; and establishing a plurality of corresponding blood vessel trees based on the plurality of connected subgraphs in the blood vessel graph. The method can analyze the shape, structure and characteristics of the blood vessel image and the connection relation between the blood vessels, and construct an accurate blood vessel tree.

Description

Vessel tree generation and lesion identification method, device, equipment and readable storage medium

Technical Field

The invention relates to the technical field of medical image processing and application, in particular to a vessel tree generation method, a lesion identification method, a lesion study and judgment method, a device, equipment and a readable storage medium.

Background

In the field of medical imaging, computer automatic extraction, modeling and analysis of blood vessel images, structures and forms are a basic problem for realizing intelligent medical treatment. For example, in the field of CT (Computed Tomography), effective information abstraction, extraction and calculation of the vascular structure in the outputted medical image can greatly improve the automatic identification rate of related lesions.

Currently, the research on the blood vessel image in the medical image mainly focuses on the following aspects: firstly, the blood vessel structure is displayed in an original image in an enhanced manner by adopting an image processing method, so that the aim of assisting medical staff in diagnosing is fulfilled; and secondly, the blood vessel image is separated and extracted from the medical image by an image and graphics method, so that the input which is more beneficial to recognition is provided for an intelligent diagnosis algorithm.

In the first aspect, the effect is mainly to enhance the display of the blood vessel image, but the blood vessel structure is not processed. In the second aspect, the blood vessel image is simply extracted, but the shape, structure, characteristics and connection relationship between blood vessel segments are not analyzed and modeled.

The above information disclosed in this background section is only for enhancement of understanding of the background of the invention and therefore it may contain information that does not constitute prior art that is already known to a person of ordinary skill in the art.

Disclosure of Invention

In view of the above, the present invention provides a vessel tree generation method, a lesion identification method, a lesion studying and judging method, an apparatus, a device and a readable storage medium.

Additional features and advantages of the invention will be set forth in the detailed description which follows, or may be learned by practice of the invention.

According to an aspect of the present invention, there is provided a medical image-based vessel tree generation method, including: acquiring a medical image, the medical image comprising a plurality of voxels; segmenting a plurality of vessel regions from the medical image, the vessel regions comprising: blood vessels and voxels with similar characteristics to blood vessels; establishing an incidence matrix of the plurality of blood vessel regions according to the adjacent relation among the plurality of blood vessel regions; respectively acquiring a blood vessel section and a blood vessel bifurcation point of each blood vessel region; constructing the blood vessel sections, the blood vessel bifurcation points and the connection relation among the blood vessel sections and the blood vessel bifurcation points which are respectively contained in each blood vessel area into a blood vessel graph with the blood vessel sections and the blood vessel bifurcation points as nodes according to the incidence matrix; establishing a plurality of corresponding vessel trees based on a plurality of connected subgraphs in the vessel map; and the nodes in the blood vessel tree are the blood vessel segments contained in the corresponding connected subgraph respectively.

According to an embodiment of the present invention, the establishing the association matrix of the plurality of blood vessel regions according to the proximity relationship between the plurality of blood vessel regions includes: for each slice image of the medical image: inserting the plurality of blood vessel regions into a spatial index structure by taking the region covered by the layer as a node; for each blood vessel region, expanding adjacent voxels to obtain an expanded region of the blood vessel region; searching an overlapped blood vessel region overlapped with the blood vessel region in the spatial index structure by taking the expanded region of the blood vessel region as a spatial query condition; assigning values to corresponding elements in the incidence matrix according to the expansion times of each blood vessel region; and obtaining the correlation matrix for the plurality of vessel regions.

According to an embodiment of the present invention, the obtaining the vessel segment and the vessel bifurcation point of each of the vessel regions respectively comprises: for each of the vessel regions: generating a plurality of corresponding three-dimensional skeletons for the blood vessel region through computational fractal theory; respectively obtaining the vessel bifurcation point from each three-dimensional skeleton, wherein the vessel bifurcation point is a voxel with 2 or more than 2 adjacent voxels around; and for each said three-dimensional skeleton: removing the vessel bifurcation point to obtain a plurality of vessel segment skeletons in the three-dimensional skeleton; respectively carrying out region growing on the plurality of vascular section skeletons to obtain a plurality of vascular sections respectively corresponding to the plurality of vascular section skeletons; and determining attributes of the plurality of vessel segments, respectively, the attributes including: length, radius, direction, voxel set of the vessel segment.

According to an embodiment of the present invention, each of the vessel trees includes a trunk node and a branch node; establishing a plurality of corresponding vessel trees based on the plurality of connected subgraphs in the vessel map comprises: selecting the blood vessel section which is closest to the heart, has the largest radius and only has one blood vessel bifurcation point adjacent to the blood vessel section as a root node of the blood vessel tree aiming at each connected subgraph; and sequentially processing all nodes in the connected subgraph from the root node, wherein the processing comprises the following steps: for each adjacent vessel bifurcation point of the node: acquiring adjacent nodes of the vessel bifurcation; selecting one of the adjacent nodes as the trunk node, and marking other adjacent nodes as the branch nodes; connecting the node with the trunk node to form a trunk edge; and connecting the node with the branch node to form a branch edge.

According to an embodiment of the present invention, performing region growing on the plurality of blood vessel segment skeletons respectively to obtain a plurality of blood vessel segments corresponding to the plurality of blood vessel segment skeletons respectively includes: construction ofA voxel incidence matrix A of the blood vessel region, wherein when two voxels are adjacent, the element in the incidence matrix A is 1; when two voxels are not adjacent, the element in the incidence matrix A is 0; constructing a vector b of each blood vessel section, and assigning points in the vector b as serial numbers corresponding to the blood vessel section frameworks; repeating the following steps until all elements in the vector b are not 0, thereby obtaining the vessel segment: 1) let g be _ij＝a_ijb_jWherein a is_ijIs an element in the incidence matrix A, b_jIs an element in the vector b; 2) b is caused to be_i’＝max(g_i1,g_i2,…,g_in) (ii) a 3) If b is_jIs equal to 0, then b is obtained_jIs equal to b_i’。

According to an aspect of the present invention, there is provided a method for identifying a lesion based on a vascular tree, comprising: and recognizing the lesion according to a predefined lesion recognition rule based on the blood vessel tree generated by the blood vessel tree generating method based on the medical image.

According to an embodiment of the invention, the lesion identification rules comprise some or all of the following rules: when the number of nodes in the blood vessel tree is smaller than a preset node number threshold value, determining that a region corresponding to the blood vessel tree has a lesion; when the radius difference values of the nodes in the blood vessel tree and the adjacent nodes are all larger than a preset first radius threshold value, determining that the region corresponding to the node has a lesion; when a breadth-first traversal strategy is adopted to traverse each node of the blood vessel tree, determining that a region corresponding to a node, the radius difference value of which is beyond a preset radius range, of two nodes visited before and after has a lesion; when the radius of a branch node of the blood vessel tree is larger than a preset second radius threshold, determining that a lesion exists in an area corresponding to the branch node; for the medical image of a bilaterally symmetric organ, when the KL distance of the radius distribution curve of the nodes in the two largest blood vessel trees on the left side and the right side of the organ is greater than a preset first distance threshold value, determining that the regions corresponding to the two blood vessel trees have lesions; and for the medical image which is an image of a left-right asymmetric organ, when the KL distance between the radius distribution curve of the nodes in the blood vessel tree and the standard radius curve is greater than a preset second distance threshold value, determining that a region corresponding to the blood vessel tree has a lesion.

According to an aspect of the present invention, there is provided a method for studying and judging a lesion region based on a blood vessel tree, including: and (3) based on the blood vessel tree generated by the blood vessel tree generation method based on any medical image, carrying out lesion study and judgment on the region to be studied and judged according to a predefined lesion study and judgment rule.

According to an embodiment of the present invention, the lesion studying and judging rule includes some or all of the following rules: when the difference value between the radius of the region to be judged and the radius of the node of the blood vessel tree where the region to be judged is located is larger than a preset first radius threshold value, determining that the region to be judged is a lesion region; when the number of nodes of the blood vessel tree where the region to be researched is located is less than a preset first node number threshold value, determining that the region to be researched is a lesion region; when the difference value between the radius of the region to be judged and the radius of all adjacent nodes of the blood vessel tree where the region to be judged is located is larger than a preset second radius threshold value, determining that the region to be judged is a lesion region; and when the radius of the node of the blood vessel tree where the region to be judged is located is larger than a third radius threshold value and the nodes of the subtree of the nodes are smaller than a second node number threshold value, determining that the region to be judged is a lesion region.

According to an aspect of the present invention, there is provided a medical image-based vessel tree generating apparatus, including: an image acquisition module for acquiring a medical image, the medical image comprising a plurality of voxels; a region segmentation module for segmenting a plurality of vessel regions from the medical image, the vessel regions comprising: blood vessels and voxels with similar characteristics to blood vessels; the matrix establishing module is used for establishing an incidence matrix of the plurality of blood vessel regions according to the adjacent relation among the plurality of blood vessel regions; the blood vessel section acquisition module is used for respectively acquiring the blood vessel section and the blood vessel bifurcation of each blood vessel region; the blood vessel map construction module is used for constructing a blood vessel map which takes the blood vessel sections and the blood vessel bifurcation points as nodes according to the incidence matrix and the connection relation among the blood vessel sections and the blood vessel bifurcation points which are respectively contained in each blood vessel region; the vessel tree establishing module is used for establishing a plurality of corresponding vessel trees based on a plurality of connected subgraphs in the vessel map; wherein each vessel tree comprises a trunk node and branch nodes.

According to an aspect of the present invention, there is provided a vessel tree based lesion recognition device, including: and the lesion identification module is used for identifying lesions according to predefined lesion identification rules on the basis of the blood vessel tree generated by the blood vessel tree generation method based on any one of the medical images.

According to an aspect of the present invention, there is provided a lesion region studying and judging device based on a blood vessel tree, including: and the lesion studying and judging module is used for studying and judging lesions of the regions to be studied and judged according to predefined lesion studying and judging rules on the basis of the blood vessel tree generated by the blood vessel tree generating method based on any one of the medical images.

According to an aspect of the present invention, there is provided a computer apparatus comprising: a memory, a processor and executable instructions stored in the memory and executable in the processor, the processor implementing any of the methods described above when executing the executable instructions.

According to an aspect of the invention, there is provided a computer-readable storage medium having stored thereon computer-executable instructions which, when executed by a processor, implement any of the methods described above.

According to the blood vessel tree generation method based on the medical image, provided by the embodiment of the invention, the shape, the structure and the characteristics of the blood vessel image and the connection relation between blood vessels can be analyzed aiming at the divided blood vessel image, and the accurate blood vessel tree is constructed. The vessel tree can be used for lesion identification, and can also be used for further studying and judging lesion areas identified by other modes so as to obtain a more accurate lesion identification result.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.

Drawings

The above and other objects, features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

Fig. 1 is a flow chart illustrating a medical image-based vessel tree generation method according to an exemplary embodiment.

Fig. 2 is a cross-sectional view of a vessel region at a layer, according to an example.

Fig. 3 and 4 show a vessel tree generated according to the above vessel tree generation method based on the scanned retinal image and the lung image, respectively.

Fig. 5 is a flow chart illustrating another medical image-based vessel tree generation method according to an exemplary embodiment.

Fig. 6 is a flowchart illustrating yet another medical image-based vessel tree generation method according to an exemplary embodiment.

Fig. 7 shows a cross-sectional view of one layer of the three-dimensional scaffold generated for the vascular region shown in fig. 2.

Fig. 8 is a schematic diagram of a vessel segment generation process according to an example shown.

Fig. 9 is a flowchart illustrating yet another medical image-based vessel tree generation method according to an exemplary embodiment.

Fig. 10 shows a process of generating a vessel tree from the vessel segments as shown in fig. 8.

FIG. 11 is a flow chart illustrating a vessel tree based lesion identification method according to an exemplary embodiment.

Fig. 12 shows a schematic view of lung lesions identified according to the identification rule.

Fig. 13 is a flow chart illustrating a method for vessel tree based lesion area study according to an exemplary embodiment.

Fig. 14 is a block diagram illustrating a medical image-based vessel tree generating apparatus according to an exemplary embodiment.

Fig. 15 is a block diagram illustrating a vessel tree based lesion recognition device according to an exemplary embodiment.

Fig. 16 is a block diagram illustrating a lesion region studying apparatus based on a blood vessel tree according to an exemplary embodiment.

Fig. 17 is a schematic structural diagram of an electronic device shown in accordance with an example embodiment.

FIG. 18 is a schematic diagram illustrating a computer-readable storage medium according to an example embodiment.

Detailed Description

Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The drawings are merely schematic illustrations of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repetitive description will be omitted.

Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to provide a thorough understanding of embodiments of the invention. One skilled in the relevant art will recognize, however, that the invention may be practiced without one or more of the specific details, or with other methods, components, devices, steps, and so forth. In other instances, well-known structures, methods, devices, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the invention.

Further, in the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise. "and/or" describes the association relationship of the associated objects, and means that there may be three relationships, for example, a and/or B, and that there may be three cases of a alone, B alone, and a and B simultaneously. The symbol "/" generally indicates that the former and latter associated objects are in an "or" relationship. The terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

Referring to fig. 1, a method 10 for generating a medical image-based vessel tree includes:

in step S102, a medical image is acquired.

The medical image may include, for example, a CT medical image, an MRI (Magnetic Resonance Imaging) medical image, an ultrasound medical image, or the like; in addition, the organs collected in the medical image may be lung, liver, brain, retina, etc., and the invention is not limited thereto.

Taking a common axial CT scan image as an example, the medical image may be a three-dimensional tensor Q in which each element Q is_i,j,kAre non-negative integers and may also be referred to as voxels or pixels. Wherein, the subscript i is 0,1,2,3, … …, which is a row index in a certain layer of the CT scan; subscript j ═ 0,1,2,3, … …, which is the column index in a slice of the CT scan; k is 0,1,2,3, … … is the slice index for CT scan.

In step S104, a plurality of blood vessel regions are segmented from the medical image.

The medical image is subjected to image preprocessing, and voxel values of non-blood vessel regions (for example, air regions of the lung in the case of a lung image) and other tissue regions except for blood vessel regions and voxels having similar features to blood vessels are set to 0, so that a plurality of blood vessel regions can be segmented from the medical image.

The vascular region includes: blood vessels and voxels with similar characteristics to blood vessels.

One blood vessel region is defined as r_x，r_xe.R, where R is the set of vessel regions segmented from the medical image. Each blood vessel region is oneA separate three-dimensional region, and each vessel region is not adjacent to other vessel regions. That is to say, the axially adjacent voxel values of each vessel region are all 0. Fig. 2 is a cross-sectional view of a vessel region at a layer, according to an example. Wherein the black voxel value is 0 and the white voxel value is greater than 0.

In step S106, a correlation matrix of the plurality of blood vessel regions is established according to the proximity relationship between the plurality of blood vessel regions.

The correlation matrix of a plurality of blood vessel regions can be denoted as M, and is used for representing the neighbor relation between the blood vessel regions, i.e. which blood vessel regions are closer to each other. Since the closer vessel regions are likely to be connected vessels in the image, these vessels are disconnected only by the decision made by thresholding or by the noise of the CT scan. The construction of the incidence matrix helps to establish possible connection relationships between the vessel regions.

In step S108, a vessel segment and a vessel bifurcation point of each vessel region are acquired, respectively.

For example, the vessel segment and vessel bifurcation included in each vessel region can be obtained by computational fractal theory (also referred to as computational morphology).

In step S110, the vessel segments, the vessel bifurcation points, and the connection relationship between the vessel segments and the vessel bifurcation points included in each vessel region are constructed into a vessel map using the vessel segments and the vessel bifurcation points as nodes according to the correlation matrix.

For example, the vessel segments, vessel bifurcation points, and the connection relationship between the vessel segments and the vessel bifurcation points included in each vessel region may be first constructed into a preliminary vessel map, and then the connection relationship between the vessel regions is added to the preliminary vessel map according to the correlation matrix M to obtain a final vessel map (e.g. may be denoted as G).

In step S112, a plurality of corresponding vessel trees are established based on the plurality of connected subgraphs in the vessel map.

And the nodes in each blood vessel tree are blood vessel sections contained in the corresponding connected subgraph.

The blood vessel map can comprise a plurality of connected subgraphs, wherein the blood vessel segments in each connected subgraph are connected with each other, and the blood vessels between the connected subgraphs are not connected. And establishing corresponding blood vessel trees based on the connected subgraphs. For example, the vessels between the left and right lungs are not connected to each other, and thus belong to different connected subgraphs, and different vessel trees can be constructed respectively.

According to the blood vessel tree generation method based on the medical image, provided by the embodiment of the invention, the shape, the structure and the characteristics of the blood vessel image and the connection relation between blood vessels can be analyzed aiming at the divided blood vessel image, and the accurate blood vessel tree is constructed. The vessel tree can be used for lesion identification, and can also be used for further studying and judging lesion areas identified by other modes so as to obtain a more accurate lesion identification result. Fig. 3 and 4 show a vessel tree generated according to the above vessel tree generation method based on the scanned retinal image and the lung image, respectively. Wherein fig. 3(a) and fig. 4(a) show the scanned retinal image and lung image, respectively, and fig. 3(b) and fig. 4(b) show the blood vessel tree corresponding to the retinal image and lung image, respectively. As can be seen from fig. 3 and 4, by this method, an accurate vessel tree can be generated based on the acquired medical images.

It should be clearly understood that the present disclosure describes how to make and use particular examples, but the principles of the present disclosure are not limited to any details of these examples. Rather, these principles can be applied to many other embodiments based on the teachings of the present disclosure.

Fig. 5 is a flow chart illustrating another medical image-based vessel tree generation method according to an exemplary embodiment. Unlike the medical image-based vessel tree generation method 10 shown in fig. 1, the medical image-based vessel tree generation method 20 shown in fig. 5 further provides a method for establishing a correlation matrix of a plurality of vessel regions according to a proximity relationship between the plurality of vessel regions.

Referring to fig. 5, the method 20 for generating a medical image-based vessel tree includes:

in step S202, for each slice of the medical image, a plurality of blood vessel regions are processed:

1. a plurality of blood vessel regions r_xTaking the region covered by the layer as a node, inserting a spatial index structure I_k。

The spatial index structure I_kFor example, the structure may be indexed by an existing Rtree structure, or may be indexed by other existing spatial index structures.

2. For each vessel region r_xExpanding the adjacent voxels to obtain the expanded region of the vessel region, which can be denoted as r_x’。

3. With the expanded region r of the vessel region_x' for spatial query condition, in spatial index structure I_kTo the vessel region r_xOverlapping vessel region r_y。

4. And assigning values to corresponding elements in the incidence matrix M according to the expansion times.

For example, the number of initialization expansion times z is 0, and r_x’＝r_x。

If the element M in the correlation matrix M is_xyIs equal to 0 or m_xyGreater than z, then z is assigned to m_xyAnd m_yx。

After the current expansion is completed, if the expansion times z do not reach a preset threshold (such as 2-6 voxels), skipping to the step 2, and repeatedly executing.

In step S204, a correlation matrix of a plurality of blood vessel regions is obtained.

After the above operations are completed for each layer, a final correlation matrix of a plurality of blood vessel regions is established.

Fig. 6 is a flowchart illustrating yet another medical image-based vessel tree generation method according to an exemplary embodiment. Unlike the medical image-based vessel tree generation method 10 shown in fig. 1, the medical image-based vessel tree generation method 30 shown in fig. 6 further provides a method for separately acquiring a vessel segment and a vessel bifurcation of each vessel region.

Referring to fig. 6, the medical image-based vessel tree generating method 30 includes: the following steps are performed for each vessel region:

in step S302, a plurality of three-dimensional skeletons are generated for the blood vessel region by calculating fractal theory.

By adopting the existing skeleton extraction (skeletonization) method in the computational fractal theory, a corresponding three-dimensional skeleton can be generated for each blood vessel region. Fig. 7 shows a cross-sectional view of one layer of the three-dimensional scaffold generated for the vascular region shown in fig. 2.

In step S304, a vessel bifurcation point is acquired from each three-dimensional skeleton, respectively.

A vessel bifurcation point is a voxel surrounded by 2 or more than 2 neighboring voxels, e.g. gray voxels as in fig. 7.

In a specific implementation, for example, the voxel value of the vessel bifurcation point may be set to-1, and it is labeled as a virtual node and added to the virtual node set S.

In step S306, for each three-dimensional skeleton: removing a vessel bifurcation point to obtain a plurality of vessel section skeletons in the three-dimensional skeleton; respectively carrying out region growth on the blood vessel section skeletons to obtain a plurality of blood vessel sections respectively corresponding to the blood vessel section skeletons; and determining attributes of the plurality of vessel segments, respectively.

Fig. 8 is a schematic diagram of a vessel segment generation process according to an example shown. Fig. 8(a) shows the respective vessel bifurcation points s of the skeleton.

After removing each vascular bifurcation s, a labeling (labeling) method of the connected region was used to obtain a skeleton of each vascular segment as shown in fig. 8 (b). The vessel segment skeleton is effectively the centerline of the vessel segment without any vessel branches.

The skeleton of each vessel segment is subjected to regional growth and reduced into a vessel segment as shown in fig. 8 (c). Specifically, voxels in adjacent vessel regions are circumferentially expanded sequentially for each vessel segment until no vessel region voxel can be expanded.

The algorithmic approach may include, for example: and carrying out algorithm design based on a Matrix-Vector Iteration framework.

A voxel association matrix a of the vessel region is defined, wherein the element in the corresponding matrix a is 1 when two voxels are adjacent, and 0 otherwise. The vessel segments of these voxels are assigned a vector b. Since the vessel segment skeleton is known and marked with a number greater than 0, the points therein are assigned on vector b as the number of the vessel segment skeleton.

The following operations were repeated:

1. let g be_ij＝a_ijb_jRepresenting the voxel i by the edge a_ijCan reach the blood vessel skeleton b_j；

2. B is caused to be_i’＝max(g_i1,g_i2,…,g_in) Which determines the vascular segment skeleton that voxel i can ultimately reach;

3. if b is_iIs equal to 0, then b is obtained_i＝b_i’。

Until all elements in vector b are not 0, an assignment is obtained, and thus region growing from the vessel segment skeleton to the vessel segment is completed.

The set of vessel segments may be denoted V, for example, with each vessel segment V ∈ V.

Then, calculating the attribute of each blood vessel segment separately, wherein the blood vessel segment attribute may include: the length, radius, direction, voxel set, etc. of the vessel segment may be denoted as { l, r, θ, C }, respectively.

Where the set of voxels to which each vessel segment v corresponds is denoted C.

The radius of the vessel section v can be calculated by equation (1):

wherein the function dist (P, P) _c) Representing the skeleton P corresponding from the point P belonging to O to the vessel segment v_cThe distance of (c).

Calculating the direction θ of each vessel segment v: the main distribution direction of all voxels C of the vessel segment is taken as the direction of the vessel segment v. The set of voxels C of the vessel segment is represented as a 3 x N matrix of voxel point coordinates, of which each three-dimensional column vector C e C is one. Therefore, the voxel point distribution can be obtained by Principal Component Analysis (PCA)Direction, i.e.

Wherein

Is the average coordinate of the voxels in C.

The length l of the vessel section v can be calculated by equation (2):

where T denotes transposition.

Fig. 9 is a flowchart illustrating yet another medical image-based vessel tree generation method according to an exemplary embodiment. Unlike the method 10 for generating a blood vessel tree based on medical image shown in fig. 1, the method 40 for generating a blood vessel tree based on medical image shown in fig. 9 further provides a method for establishing a plurality of corresponding blood vessel trees based on a plurality of connected subgraphs in the blood vessel map.

Referring to fig. 9, the medical image-based vessel tree generating method 40 includes:

in step S402, for each connected subgraph, a vessel segment which is closest to the heart, has the largest radius, and has only one vessel bifurcation point adjacent to the vessel segment is selected as a root node of the vessel tree.

For example, the root node may be denoted as v_c. Fig. 10 shows a process of generating a vessel tree from the vessel segments as shown in fig. 8. Wherein fig. 10(a) is a schematic view of a plurality of vessel segments and vessel bifurcations, and fig. 10(b) is a schematic view of one communicating subgraph of a vessel graph constructed from the vessel segments and vessel bifurcations in fig. 10 (a). Fig. 10(c) shows a schematic diagram of a vessel tree generated from the connected subgraph shown in fig. 10 (b).

As shown in FIG. 10(b), the vessel segment v₀The selection rule of the vessel tree root nodes is met.

In step S404, sequentially processing all nodes in the connected subgraph from the root node, including: for each adjacent vessel bifurcation point of the node:

1. adjacent nodes of the vessel bifurcation are obtained.

To process the root node v as in FIG. 10(b)₀For example, the vascular bifurcation s₀Are respectively blood vessel sections v₁And v₄,。

2. And selecting one of the adjacent nodes as a trunk node of the blood vessel tree, and marking other adjacent nodes as branch nodes.

Selecting one of the neighboring nodes as a trunk node of the vessel tree may include, for example: selecting a trunk node according to the formula (3):

wherein v is_τThe node of the backbone is represented by a representation,

A set of neighboring nodes representing the vessel bifurcation point. 2-5 voxels, representing the upper bound of the difference in radius between vascular nodes

3. The node is connected to the backbone node, forming and marking it as the backbone edge τ.

4. The node is connected with the branch node, and the node is formed and marked as a branch edge b.

The constructed vessel tree is shown in fig. 10 (c).

Referring to fig. 11, a vessel tree based lesion recognition method 50 includes:

in step S502, lesion recognition is performed according to a predefined lesion recognition rule based on the vessel tree generated according to any one of the above-described medical image-based vessel tree generation methods 10 to 40.

In some embodiments, the lesion identification rules include some or all of the following rules:

and when the number of the nodes in the blood vessel tree is smaller than a preset node number threshold value, determining that the region corresponding to the blood vessel tree has the lesion.

For example, if the vessel tree has only isolated root nodes, or the nodes are very few, that is, the depth of the root of the vessel tree is very small, it is determined that the region corresponding to the vessel tree has a lesion. The node number threshold may be set according to actual requirements when applied.

And when the radius difference values of the nodes in the blood vessel tree and the adjacent nodes are all larger than a preset first radius threshold value, determining that the region corresponding to the node has a lesion.

The first radius threshold may be set to 1-5 voxels, for example, but the invention is not limited thereto.

When a breadth-first traversal strategy is adopted to traverse each node of the blood vessel tree, determining that the region corresponding to the node with the radius difference value of the two nodes visited before and after falling outside the preset radius range has a lesion.

The radius range may be set to [ -1,2], for example, but the present invention is not limited thereto.

And when the radius of the branch node of the blood vessel tree is larger than a preset second radius threshold value, determining that the region corresponding to the branch node has a lesion.

The second radius threshold may be set to 1-5, for example, but the invention is not limited thereto.

And for the medical image of the bilaterally symmetric organ, determining that the region corresponding to the two vessel trees has the lesion when the KL distance of the radius distribution curves of the nodes in the two largest vessel trees on the left and right sides of the organ is greater than a preset first distance threshold.

The KL distance is an abbreviation for Kullback-Leibler difference (Kullback-Leibler bias), also called Relative Entropy (Relative Entropy). It measures the difference between two probability distributions in the same event space.

Specifically, assuming that the radius distribution vectors of the nodes in the two vessel trees are p1 and p2, the KL distances thereof are calculated according to formula (4):

wherein p1 and p2 are normalized vectors, p1(x) and p2(x) are radius distribution functions of nodes in two vessel trees, namely the number proportion of vessel segments with radius x is p1(x) and p2 (x).

Fig. 12 shows a schematic view of lung lesions identified according to the identification rule. As shown in fig. 12(a), the KL distance between the distribution curves of the radii of the nodes in the two corresponding vessel trees of the left and right lungs is greater than the first distance threshold, and corresponding to the CT scan taken for each slice, a large slice of lesion as shown in fig. 12(b) is found and can be seen in many slices.

And for the medical image which is an image of a left-right asymmetric organ, when the KL distance between the radius distribution curve of the nodes in the blood vessel tree and the standard radius curve is greater than a preset second distance threshold value, determining that the region corresponding to the blood vessel tree has a lesion.

The standard radius curve can be obtained, for example, by counting the distribution curves of node radii in the normalized vessel tree of all normal people.

Normalization method for (x, y, z) ∈ Q_lThe following transformation is performed to obtain:

x’＝(x-x_min)/(x_max-x_min)

y’＝(y-y_min)/(y_max-x_min)

z’＝(z-z_min)/(z_max-z_min)

Ql denotes an organ segmentation region tensor of the three-dimensional tensor Q of the CT image, and x _ min, and x _ max denotes a minimum abscissa and a maximum abscissa of the Ql, which are not 0 voxel coordinates. The other coordinates are the same.

The first distance threshold and the second distance threshold may be set according to actual requirements when applied.

According to the lesion identification method based on the blood vessel tree, the lesion in the medical image can be accurately identified according to the predefined lesion identification rule based on the constructed blood vessel tree.

Referring to fig. 13, the method 60 for determining a lesion region based on a blood vessel tree includes:

in step S602, based on the blood vessel tree generated according to any one of the above blood vessel tree generation methods 10 to 40 based on medical images, lesion study and judgment are performed on a region to be studied and judged according to predefined lesion study and judgment rules.

In some embodiments, the lesion study rules include some or all of the following rules:

and when the difference value between the radius of the region to be judged and the radius of the node of the blood vessel tree where the region to be judged is located is larger than a preset first radius threshold value, determining the region to be judged as a lesion region.

The first radius threshold may be set to 1-5, for example, but the invention is not limited thereto.

And when the number of nodes of the blood vessel tree in which the region to be judged is located is less than a preset first node number threshold value, determining the region to be judged as a lesion region.

And when the difference value between the radius of the region to be judged and the radius of all adjacent nodes of the node of the blood vessel tree where the region to be judged is located is larger than a preset second radius threshold value, determining that the region to be judged is a lesion region.

And when the radius of the node of the blood vessel tree where the region to be judged is located is larger than a third radius threshold (for example, 0.5-2) and the nodes of the subtree of the nodes are smaller than a second node number threshold, determining that the region to be judged is a lesion region.

It should be noted that, when the above thresholds are applied, they may be set according to actual requirements, and the present invention is not limited thereto.

According to the method for studying and judging the lesion area based on the blood vessel tree, which is provided by the embodiment of the invention, the constructed blood vessel tree is based on the predefined lesion studying and judging rule, and the identified lesion area can be further studied and judged so as to improve the accuracy of lesion identification.

Those skilled in the art will appreciate that all or part of the steps implementing the above embodiments are implemented as computer programs executed by a CPU. The computer program, when executed by the CPU, performs the functions defined by the method provided by the present invention. The program may be stored in a computer readable storage medium, which may be a read-only memory, a magnetic or optical disk, or the like.

Furthermore, it should be noted that the above-mentioned figures are only schematic illustrations of the processes involved in the method according to exemplary embodiments of the invention, and are not intended to be limiting. It will be readily understood that the processes shown in the above figures are not intended to indicate or limit the chronological order of the processes. In addition, it is also readily understood that these processes may be performed synchronously or asynchronously, e.g., in multiple modules.

The following are embodiments of the apparatus of the present invention that may be used to perform embodiments of the method of the present invention. For details which are not disclosed in the embodiments of the apparatus of the present invention, reference is made to the embodiments of the method of the present invention.

Referring to fig. 14, the medical image-based blood vessel tree generating apparatus 70 includes: an image acquisition module 702, a region segmentation module 704, a matrix establishment module 706, a vessel segment acquisition module 708, a vessel map construction module 710, and a vessel tree establishment module 712.

Therein, the image acquisition module 702 is configured to acquire a medical image, which includes a plurality of voxels.

The region segmentation module 704 is configured to segment a plurality of blood vessel regions from the medical image, where the blood vessel regions include: blood vessels and voxels with similar characteristics to blood vessels.

The matrix establishing module 706 is configured to establish a correlation matrix of the plurality of blood vessel regions according to a proximity relationship between the plurality of blood vessel regions.

The vessel segment acquiring module 708 is configured to acquire a vessel segment and a vessel bifurcation point of each vessel region respectively.

The vessel map constructing module 710 is configured to construct a vessel map using the vessel segments and the vessel bifurcation points as nodes according to the association matrix and the vessel segments, the vessel bifurcation points and the connection relationship among the vessel segments and the vessel bifurcation points included in each vessel region.

The vessel tree establishing module 712 is configured to establish a plurality of corresponding vessel trees based on the plurality of connected subgraphs in the vessel map.

Wherein each vessel tree comprises a trunk node and branch nodes.

In some embodiments, the matrix building module 706 includes: a first establishing unit and a second establishing unit. Wherein the first establishing unit is used for aiming at each layer image of the medical image: inserting a plurality of blood vessel regions into the spatial index structure by taking the region covered by the layer as a node; for each blood vessel region, expanding adjacent voxels to obtain an expanded region of the blood vessel region; searching an overlapped blood vessel region overlapped with the blood vessel region in a spatial index structure by taking the expanded region of the blood vessel region as a spatial query condition; and assigning values to corresponding elements in the incidence matrix according to the expansion times of each blood vessel region. The second establishing unit is used for obtaining a correlation matrix of a plurality of blood vessel regions.

In some embodiments, vessel segment acquisition module 708 includes: skeleton generation unit, bifurcation point remove unit and blood vessel section acquisition unit. The skeleton generation unit is used for generating a plurality of corresponding three-dimensional skeletons for the blood vessel region through computational fractal theory. The bifurcation point removing unit is used for respectively obtaining a blood vessel bifurcation point from each three-dimensional skeleton, wherein the blood vessel bifurcation point is a voxel around which 2 or more than 2 adjacent voxels surround. The vessel segment obtaining unit is configured to, for each three-dimensional skeleton: removing a vessel bifurcation point to obtain a plurality of vessel section skeletons in the three-dimensional skeleton; respectively carrying out region growth on the blood vessel section skeletons to obtain a plurality of blood vessel sections respectively corresponding to the blood vessel section skeletons; and determining attributes of the plurality of vessel segments respectively, the attributes including: length, radius, direction, voxel set of vessel segment.

In some embodiments, the vessel segment obtaining unit further comprises: the device comprises a matrix building subunit, a vector building subunit and a region growing subunit. The matrix construction subunit is used for constructing a voxel association matrix A of the blood vessel region, wherein when two voxels are adjacent, the element in the association matrix A is 1; when two voxels are not adjacent, the element in the incidence matrix A is 0; the vector construction subunit is used for constructing a vector b of each blood vessel section, and points in the vector b are assigned as numbers corresponding to the blood vessel section frameworks; the region growing subunit is configured to repeat the following steps until all elements in the vector b are not 0, thereby obtaining the vessel segment:

1) Let g be_ij＝a_ijb_jWherein a is_ijIs an element in the incidence matrix A, b_jIs an element in the vector b;

2) b is caused to be_i’＝max(g_i1,g_i2,…,g_in)；

3) If b is_jIs equal to 0, then b is obtained_jIs equal to b_i’。

In some embodiments, each vessel tree includes a trunk node and branch nodes. The vessel tree building module 712 includes: the first establishing unit is used for selecting a blood vessel section which is closest to the heart, has the largest radius and only has one blood vessel bifurcation point adjacent to the blood vessel section as a root node of the blood vessel tree aiming at each connected subgraph; and starting from the root node, sequentially processing all nodes in the connected subgraph, wherein the processing comprises the following steps: for each adjacent vessel bifurcation point of the node: acquiring adjacent nodes of the vessel bifurcation; selecting one of the adjacent nodes as a trunk node, and marking other adjacent nodes as branch nodes; connecting the node with the trunk node to form a trunk edge; and connecting the node and the branch node to form a branch edge.

According to the blood vessel tree generator device based on the medical image, provided by the embodiment of the invention, the shape, the structure and the characteristics of the blood vessel image and the connection relation between blood vessels can be analyzed aiming at the divided blood vessel image, and the accurate blood vessel tree can be constructed. The vessel tree can be used for lesion identification, and can also be used for further studying and judging lesion areas identified by other modes so as to obtain a more accurate lesion identification result.

Referring to fig. 15, the blood vessel tree-based lesion recognition device 80 includes: a lesion identification module 802, configured to identify a lesion based on a blood vessel tree generated according to any one of the above-mentioned medical image-based blood vessel tree generation methods 10 to 40 according to a predefined lesion identification rule.

when the difference value between the radius of the area to be judged and the radius of the node of the blood vessel tree where the area to be judged is located is larger than a preset first radius threshold value, determining the area to be judged as a lesion area;

when the number of nodes of a blood vessel tree where the region to be researched is located is less than a preset first node number threshold value, determining the region to be researched as a lesion region;

when the difference value between the radius of the region to be judged and the radius of all adjacent nodes of the node of the blood vessel tree where the region to be judged is located is larger than a preset second radius threshold value, determining the region to be judged as a lesion region;

and when the radius of the node of the blood vessel tree where the region to be judged is located is larger than the third radius threshold value and the nodes of the subtree of the nodes are smaller than the second node number threshold value, determining that the region to be judged is a lesion region.

According to the lesion identification device based on the blood vessel tree provided by the embodiment of the invention, based on the constructed blood vessel tree, according to the predefined lesion identification rule, the lesion in the medical image can be accurately identified.

Referring to fig. 16, the lesion region study device 90 based on the blood vessel tree includes: a lesion studying and judging module 902, configured to study and judge a lesion in a region to be studied and judged according to a predefined lesion studying and judging rule based on a vessel tree generated according to any one of the vessel tree generation methods 10 to 40 based on the medical image.

According to the pathological change region studying and judging device based on the blood vessel tree, the established blood vessel tree is based on, and the identified pathological change region can be further studied and judged according to the predefined pathological change studying and judging rule, so that the accuracy of pathological change identification is improved.

It is noted that the block diagrams shown in the above figures are functional entities and do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices.

Fig. 17 is a schematic structural diagram of an electronic device shown in accordance with an example embodiment. It should be noted that the electronic device shown in fig. 17 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiment of the present invention.

As shown in fig. 17, the electronic device 800 is embodied in the form of a general-purpose computer device. The components of the electronic device 800 include: at least one Central Processing Unit (CPU)801, which may perform various appropriate actions and processes according to program code stored in a Read Only Memory (ROM)802 or loaded from at least one storage unit 808 into a Random Access Memory (RAM) 803.

In particular, according to an embodiment of the present invention, the program code may be executed by the central processing unit 801, such that the central processing unit 801 performs the steps according to various exemplary embodiments of the present invention described in the above-mentioned method embodiment section of the present specification. For example, the central processing unit 801 may perform the steps as shown in fig. 1, 5, 6, 9, 11 or 14.

In the RAM 803, various programs and data necessary for the operation of the electronic apparatus 800 are also stored. The CPU 801, ROM 802, and RAM 803 are connected to each other via a bus 804. An input/output (I/O) interface 805 is also connected to bus 804.

The following components are connected to the I/O interface 805: an input unit 806 including a keyboard, a mouse, and the like; an output unit 807 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage unit 808 including a hard disk and the like; and a communication unit 809 including a network interface card such as a LAN card, a modem, or the like. The communication unit 809 performs communication processing via a network such as the internet. A drive 810 is also connected to the I/O interface 805 as necessary. A removable medium 811 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 810 as necessary, so that a computer program read out therefrom is mounted on the storage unit 808 as necessary.

Referring to fig. 18, a program product 900 configured to implement the above method according to an embodiment of the present invention is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited in this regard and, in the present document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

The computer readable medium carries one or more programs which, when executed by a device, cause the computer readable medium to implement the functions as shown in fig. 1, fig. 5, fig. 6, fig. 9, fig. 11, or fig. 14.

Exemplary embodiments of the present invention are specifically illustrated and described above. It is to be understood that the invention is not limited to the precise construction, arrangements, or instrumentalities described herein; on the contrary, the invention is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A method for generating a blood vessel tree based on medical images is characterized by comprising the following steps:

acquiring a medical image, the medical image comprising a plurality of voxels;

segmenting a plurality of vessel regions from the medical image, the vessel regions comprising: blood vessels and voxels with similar characteristics to blood vessels;

establishing an incidence matrix of the plurality of blood vessel regions according to the adjacent relation among the plurality of blood vessel regions;

respectively acquiring a blood vessel section and a blood vessel bifurcation point of each blood vessel region;

constructing the blood vessel sections, the blood vessel bifurcation points and the connection relation among the blood vessel sections and the blood vessel bifurcation points which are respectively contained in each blood vessel area into a blood vessel graph with the blood vessel sections and the blood vessel bifurcation points as nodes according to the incidence matrix; and

establishing a plurality of corresponding vessel trees based on a plurality of connected subgraphs in the vessel map;

and the nodes in the blood vessel tree are the blood vessel segments contained in the corresponding connected subgraph respectively.

2. The method of claim 1, wherein establishing the correlation matrix of the plurality of vessel regions according to the proximity relationship between the plurality of vessel regions comprises:

for each slice image of the medical image: inserting the plurality of blood vessel regions into a spatial index structure by taking the region covered by the layer as a node; for each blood vessel region, expanding adjacent voxels to obtain an expanded region of the blood vessel region; searching an overlapped blood vessel region overlapped with the blood vessel region in the spatial index structure by taking the expanded region of the blood vessel region as a spatial query condition; assigning values to corresponding elements in the incidence matrix according to the expansion times of each blood vessel region; and

Obtaining the correlation matrix for the plurality of vessel regions.

3. The method of claim 1, wherein separately obtaining the vessel segment and vessel bifurcation of each of the vessel regions comprises: for each of the vessel regions:

generating a plurality of corresponding three-dimensional skeletons for the blood vessel region through computational fractal theory;

respectively obtaining the vessel bifurcation point from each three-dimensional skeleton, wherein the vessel bifurcation point is a voxel with 2 or more than 2 adjacent voxels around; and

for each of the three-dimensional skeletons: removing the vessel bifurcation point to obtain a plurality of vessel segment skeletons in the three-dimensional skeleton; respectively carrying out region growing on the plurality of vascular section skeletons to obtain a plurality of vascular sections respectively corresponding to the plurality of vascular section skeletons; and determining attributes of the plurality of vessel segments, respectively, the attributes including: length, radius, direction, voxel set of the vessel segment.

4. The method of claim 3, wherein the region growing the plurality of vessel segment skeletons, respectively, to obtain a plurality of vessel segments corresponding to the plurality of vessel segment skeletons, respectively, comprises:

Constructing a voxel incidence matrix A of the blood vessel region, wherein when two voxels are adjacent, the element in the incidence matrix A is 1; when two voxels are not adjacent, the element in the incidence matrix A is 0;

constructing a vector b of each blood vessel section, and assigning points in the vector b as serial numbers corresponding to the blood vessel section frameworks;

repeating the following steps until all elements in the vector b are not 0, thereby obtaining the vessel segment:

2) b is caused to be_i’＝max(g_i1,g_i2,…,g_in)；

3) If b is_jIs equal to 0, then b is obtained_jIs equal to b_i’。

5. The method according to claim 1, wherein each of the vessel trees comprises a trunk node and a branch node; establishing a plurality of corresponding vessel trees based on the plurality of connected subgraphs in the vessel map comprises: selecting the blood vessel section which is closest to the heart, has the largest radius and only has one blood vessel bifurcation point adjacent to the blood vessel section as a root node of the blood vessel tree aiming at each connected subgraph; and sequentially processing all nodes in the connected subgraph from the root node, wherein the processing comprises the following steps:

for each adjacent vessel bifurcation point of the node: acquiring adjacent nodes of the vessel bifurcation; selecting one of the adjacent nodes as the trunk node, and marking other adjacent nodes as the branch nodes; connecting the node with the trunk node to form a trunk edge; and connecting the node with the branch node to form a branch edge.

6. A lesion identification method based on a blood vessel tree is characterized by comprising the following steps:

lesion recognition is performed according to predefined lesion recognition rules based on the vessel tree generated by the medical image-based vessel tree generation method according to any one of claims 1 to 5.

7. The method of claim 6, wherein the lesion identification rules include some or all of the following rules:

when the number of nodes in the blood vessel tree is smaller than a preset node number threshold value, determining that a region corresponding to the blood vessel tree has a lesion;

when the radius difference values of the nodes in the blood vessel tree and the adjacent nodes are all larger than a preset first radius threshold value, determining that the region corresponding to the node has a lesion;

when a breadth-first traversal strategy is adopted to traverse each node of the blood vessel tree, determining that a region corresponding to a node, the radius difference value of which is beyond a preset radius range, of two nodes visited before and after has a lesion;

when the radius of a branch node of the blood vessel tree is larger than a preset second radius threshold, determining that a lesion exists in an area corresponding to the branch node;

for the medical image of a bilaterally symmetric organ, when the KL distance of the radius distribution curve of the nodes in the two largest blood vessel trees on the left side and the right side of the organ is greater than a preset first distance threshold value, determining that the regions corresponding to the two blood vessel trees have lesions;

And for the medical image which is an image of a left-right asymmetric organ, when the KL distance between the radius distribution curve of the nodes in the blood vessel tree and the standard radius curve is greater than a preset second distance threshold value, determining that a region corresponding to the blood vessel tree has a lesion.

8. A method for studying and judging a lesion area based on a blood vessel tree is characterized by comprising the following steps:

lesion study and judgment are performed on the region to be studied and judged according to predefined lesion study and judgment rules based on the vessel tree generated by the vessel tree generation method based on the medical image according to any one of claims 1 to 5.

9. The method of claim 8, wherein the lesion adjudication rules include some or all of the following rules:

when the difference value between the radius of the region to be judged and the radius of the node of the blood vessel tree where the region to be judged is located is larger than a preset first radius threshold value, determining that the region to be judged is a lesion region;

when the number of nodes of the blood vessel tree where the region to be researched is located is less than a preset first node number threshold value, determining that the region to be researched is a lesion region;

when the difference value between the radius of the region to be judged and the radius of all adjacent nodes of the blood vessel tree where the region to be judged is located is larger than a preset second radius threshold value, determining that the region to be judged is a lesion region;

And when the radius of the node of the blood vessel tree where the region to be judged is located is larger than a third radius threshold value and the nodes of the subtree of the nodes are smaller than a second node number threshold value, determining that the region to be judged is a lesion region.

10. A medical image-based vessel tree generating apparatus, comprising:

an image acquisition module for acquiring a medical image, the medical image comprising a plurality of voxels;

a region segmentation module for segmenting a plurality of vessel regions from the medical image, the vessel regions comprising: blood vessels and voxels with similar characteristics to blood vessels;

the matrix establishing module is used for establishing an incidence matrix of the plurality of blood vessel regions according to the adjacent relation among the plurality of blood vessel regions;

the blood vessel section acquisition module is used for respectively acquiring the blood vessel section and the blood vessel bifurcation of each blood vessel region;

a vessel map construction module, configured to construct, according to the incidence matrix, a vessel map using the vessel segments and the vessel bifurcation points as nodes from the vessel segments, the vessel bifurcation points, and connection relationships among the vessel segments and the vessel bifurcation points included in each of the vessel regions; and

the vessel tree establishing module is used for establishing a plurality of corresponding vessel trees based on a plurality of connected subgraphs in the vessel map;

Wherein each vessel tree comprises a trunk node and branch nodes.

11. A vascular tree-based lesion recognition device, comprising:

a lesion identification module for performing lesion identification according to predefined lesion identification rules based on the vessel tree generated by the medical image-based vessel tree generation method according to any one of claims 1 to 5.

12. A lesion region studying and judging device based on a blood vessel tree is characterized by comprising:

a lesion studying and judging module, configured to study and judge a lesion of a region to be studied and judged according to predefined lesion studying and judging rules based on the blood vessel tree generated by the medical image-based blood vessel tree generating method according to any one of claims 1 to 5.

13. A computer device, comprising: memory, processor and executable instructions stored in the memory and executable in the processor, characterized in that the processor implements the method according to any of claims 1-9 when executing the executable instructions.

14. A computer-readable storage medium having computer-executable instructions stored thereon, wherein the executable instructions, when executed by a processor, implement the method of any of claims 1-9.