Blood vessel image segmentation method and device, electronic equipment and storage medium

Abstract

The application provides a blood vessel image segmentation method and device, electronic equipment and storage medium. The method comprises the following steps: acquiring an initial blood vessel image, determining a coronary rough segmentation image and a coronary skeleton image of the initial blood vessel image respectively; determining each skeleton point on a skeleton line in the coronary skeleton image, and determining a skeleton branch point in each skeleton point based on the degree of each skeleton point; identifying a vein branch image in the initial blood vessel image based on the branch point position of the skeleton branch point in the coronary rough segmentation image / the coronary skeleton image; and determining a coronary segmentation image based on the coronary rough segmentation image and the vein branch image. Through the technical scheme disclosed by the application, the problem of low accuracy of coronary segmentation of blood vessels in the prior art is solved, and the accuracy of coronary segmentation is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical image processing, and in particular to a blood vessel image segmentation method and device, an electronic device, and a storage medium. BACKGROUND

[0002] With the improvement of imaging speed and scanning accuracy of CT (Computed Tomography, i.e. electronic computed tomography) equipment, CT medical images have been widely used in heart examination and plaque diagnosis. Heart coronary artery segmentation based on CT medical images is widely used, which can extract the outline of the coronary artery and plaque in the lumen, facilitate the observation of stenosis, calcification and plaque by doctors, and provide a basis for early prevention and diagnosis of cardiovascular diseases by doctors.

[0003] At present, in the process of heart coronary artery segmentation, due to the similar image features (close distance or even adhesion, close CT value, etc.) of veins and coronary arteries, veins are easily misidentified, and the accuracy of heart coronary artery segmentation is low. SUMMARY

[0004] The present application provides a blood vessel image segmentation method, device, electronic device and storage medium, which removes the vein branches in the coronary artery rough segmentation result, to solve the problem of low accuracy of coronary artery segmentation of blood vessels in the prior art, and to improve the accuracy of coronary artery segmentation.

[0005] In a first aspect, an embodiment of the present application provides a blood vessel image segmentation method, characterized in that it comprises:

[0006] An initial blood vessel image is obtained, and a coronary artery rough segmentation image and a coronary artery skeleton image of the initial blood vessel image are determined respectively;

[0007] Each skeleton point on the skeleton line in the coronary artery skeleton image is determined, and a skeleton branch point in each skeleton point is determined based on the degree of each skeleton point;

[0008] Based on the branch point position of the skeleton branch point in the coronary artery rough segmentation image / the coronary artery skeleton image, a vein branch image in the initial blood vessel image is identified;

[0009] A coronary artery segmentation image is determined based on the coronary artery rough segmentation image and the vein branch image.

[0010] Optionally, the coronary artery rough segmentation image and the coronary artery skeleton image of the initial blood vessel image are determined respectively, comprising:

[0011] The initial blood vessel image is input into a pre-trained coronary artery rough segmentation model to obtain a coronary artery rough segmentation image output by the coronary artery rough segmentation model;

[0012] The initial blood vessel image is input into a pre-trained skeleton segmentation model to obtain the coronary artery skeleton image output by the skeleton segmentation model.

[0013] Optionally, determining the skeleton branch points among the skeleton points based on the degree of each skeleton point includes:

[0014] For any skeleton point, determine each neighboring point of the current skeleton point, and determine the degree of the current skeleton point based on the skeleton points contained in the neighboring points;

[0015] Obtain a preset degree threshold, and determine the skeleton branch points in each skeleton point based on the degree of the skeleton point and the preset degree threshold.

[0016] Optionally, determining the skeleton branch points among the skeleton points based on the degree of each skeleton point further includes:

[0017] Based on the degree of each skeleton point, determine the skeleton endpoints and multiple candidate branch points in the skeleton points, wherein the skeleton endpoints include the root endpoint and multiple terminal branch endpoints;

[0018] Determine the endpoint number of each of the last branch endpoints, and determine the skeleton connection path between the root endpoint and any two adjacent last branch endpoints. Based on the skeleton connection path, filter each candidate branch point to obtain the skeleton branch points in the candidate branch point set.

[0019] Optionally, filtering at least one candidate branch point based on the skeleton connectivity path includes:

[0020] Determine each candidate branch point that is repeatedly traversed in each skeleton connected path, and determine the skeleton distance between each candidate branch point and the root endpoint respectively;

[0021] Based on the comparison results of each skeleton distance, skeleton branch points among the candidate branch points are determined.

[0022] Optionally, identifying the venous branch image in the initial vascular image based on the branch point position of the skeleton branch point in the coarse segmentation image of the coronary artery includes:

[0023] The branch point of the skeleton is determined in the coarse segmentation image of the coronary artery, and the coronary artery branch region in the coarse segmentation image of the coronary artery is determined based on the branch point segmentation position;

[0024] A pre-trained vein branch recognition model is obtained, and vein branches are identified in the coronary artery branch region based on the vein branch recognition model to obtain a vein branch image.

[0025] Optionally, identifying the venous branch image in the initial vascular image based on the branch point position of the skeleton branch point in the coronary artery skeleton image includes:

[0026] Determine the location of the branch point skeleton in the coronary artery skeleton image, and determine the skeleton branch region in the coronary artery skeleton image based on the location of the branch point skeleton.

[0027] A pre-trained vein branch skeleton recognition model is obtained, and vein branch skeleton recognition is performed on the skeleton branch region based on the vein branch skeleton recognition model to obtain the vein branch skeleton;

[0028] Obtain the branch expansion parameters of the vein branch skeleton, and perform expansion processing on the vein branch skeleton based on the branch expansion parameters to obtain the vein branch image corresponding to the vein branch skeleton.

[0029] Secondly, embodiments of the present invention also provide a blood vessel image segmentation apparatus, characterized in that it comprises:

[0030] The image determination module is used to acquire an initial blood vessel image and determine the coarse segmentation image of the coronary artery and the coronary artery skeleton image of the initial blood vessel image, respectively.

[0031] The skeleton branch point determination module is used to determine each skeleton point on the skeleton line in the coronary skeleton image, and to determine each skeleton endpoint and candidate branch point set in each skeleton point based on the degree of each skeleton point, and to determine the skeleton branch point in the candidate branch point set based on each skeleton endpoint.

[0032] The vein branch image determination module is used to identify vein branch images in the initial blood vessel image based on the branch point positions of the skeleton branch points in the coarse segmentation image of the coarse ...

[0033] The coronary artery segmentation image determination module is used to determine the coronary artery segmentation image of the initial blood vessel image based on the coarse coronary artery segmentation image and the venous branch image.

[0034] Thirdly, embodiments of the present invention also provide an electronic device, comprising:

[0035] At least one processor; and

[0036] A memory communicatively connected to the at least one processor; wherein,

[0037] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the blood vessel image segmentation method according to any embodiment of the present invention.

[0038] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions that are used to cause a processor to execute the blood vessel image segmentation method according to any embodiment of the present invention.

[0039] The technical solution provided by this invention involves acquiring an initial vascular image, determining a coarse coarse segmentation image and a coarse skeleton image of the initial vascular image, determining each skeleton point on the skeleton line in the coarse skeleton image, and determining skeleton branch points within each skeleton point based on the degree of each skeleton point, identifying venous branch images in the initial vascular image based on the branch point positions of the skeleton branch points in the coarse coarse segmentation image / coronary skeleton image, and determining a coarse segmentation image based on the coarse coarse segmentation image and the venous branch images. This technical solution processes the vascular image to be segmented to obtain each endpoint in the blood vessel, determines candidate branch points based on each endpoint, further determines each branch point, identifies the blood vessels between each branch point to obtain venous branches, and then optimizes the coarse coarse segmentation result based on the identified venous branches to obtain the final coarse coarse segmentation result, thereby improving the accuracy of coarse coarse segmentation.

[0040] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of a blood vessel image segmentation method provided according to an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of a coarse segmentation image of a coronary artery provided according to an embodiment of the present invention;

[0044] Figure 3 This is a schematic diagram of a coronary artery skeleton image provided according to an embodiment of the present invention;

[0045] Figure 4 This is a schematic diagram of a skeleton point provided according to an embodiment of the present invention;

[0046] Figure 5 This is a schematic diagram of another skeleton point provided according to an embodiment of the present invention;

[0047] Figure 6 This is a schematic diagram of another coarse segmentation image of the coronary artery provided according to an embodiment of the present invention;

[0048] Figure 7 This is a schematic diagram of a skeleton branch point provided according to an embodiment of the present invention;

[0049] Figure 8 This is a schematic diagram of a coronary artery branching region provided according to an embodiment of the present invention;

[0050] Figure 9 This is a schematic diagram of a vein branch image provided according to an embodiment of the present invention;

[0051] Figure 10 This is a schematic diagram of a skeleton branch region provided according to an embodiment of the present invention;

[0052] Figure 11 This is a schematic diagram of a venous branch skeleton provided according to an embodiment of the present invention;

[0053] Figure 12 This is a schematic diagram of a coronary artery segmentation image provided according to an embodiment of the present invention;

[0054] Figure 13 This is a schematic diagram of the structure of a blood vessel image segmentation device according to an embodiment of the present invention;

[0055] Figure 14 This is a schematic diagram of the structure of an electronic device that implements the blood vessel image segmentation method of this invention. Detailed Implementation

[0056] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0057] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in sequences other than those illustrated or described herein.

[0058] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0059] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0060] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0061] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0062] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0063] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0064] Figure 1 The flowchart of a blood vessel image segmentation method is provided in Embodiment 1 of the present invention. This embodiment is applicable to the case of accurate segmentation of coronary arteries in blood vessels.

[0065] In existing technologies, segmentation of vascular images often results in inaccurate coronary artery segmentation due to the inclusion of intersecting venous branches. To address this issue, this embodiment provides an image segmentation method. This method processes the vascular image to be segmented to obtain the endpoints of the vessels, determines candidate branch points based on these endpoints, and further identifies each branch point. The methods then identify the vessels between each branch point to obtain the venous branches, and finally optimize the coronary artery segmentation results based on the identified venous branches, thereby improving the accuracy of coronary artery segmentation.

[0066] This method can be executed by a blood vessel image segmentation device, which can be implemented in hardware and / or software and can be configured in a smart terminal or cloud server. For example... Figure 1 As shown, the method includes:

[0067] S110. Obtain the initial blood vessel image, and determine the coarse segmentation image of the coronary artery and the coronary artery skeleton image of the initial blood vessel image respectively.

[0068] In this embodiment of the invention, the initial vascular image can be understood as a vascular image obtained by scanning the blood vessels surrounding the heart. The initial vascular image includes images of veins that collect blood returning to the heart and images of coronary arteries that supply blood from the heart to the whole body. The coarse coronary artery segmentation image is the result of coarse coronary artery segmentation from the initial vascular image. This coarse coronary artery segmentation result may contain misidentified venous branches, so venous branch identification is required to obtain accurate coronary artery segmentation results. The coronary artery skeleton image can be understood as an image containing the coronary artery skeleton lines obtained by refining the results of the coarse coronary artery segmentation. The coronary artery skeleton lines contain information about each branch in the coarse coronary artery segmentation result.

[0069] Specifically, the method for obtaining initial vascular images can be to read image data from a local database or a cloud server database to obtain the original initial vascular image, or it can be to obtain the initial vascular image by scanning the heart area of ​​the scanned object using a medical scanning device. This embodiment does not limit the acquisition method. Optionally, the medical scanning device can be, but is not limited to, a CT device, a PET (Positron Emission Computed Tomography)-CT device, and an MRI (Magnetic Resonance Imaging) device. The corresponding initial vascular image obtained can include, but is not limited to, at least one of CT image data, PET image data, and MRI image data.

[0070] Based on the initial vascular image, the technical solution of this embodiment can acquire a pre-trained coarse coarse segmentation model for the coarse segmentation of the coarse coarse blood vessels. The initial vascular image is then input into this model to obtain a coarsely segmented image of the coarse blood vessels output by the model. Optionally, this embodiment can also use other existing segmentation techniques to perform coarse segmentation processing on the initial vascular image to obtain a coarsely segmented image of the coarse blood vessels. This embodiment does not limit the method for obtaining the coarsely segmented image of the coarse blood vessels. See also the exemplary embodiments. Figure 2 , Figure 2 The coarse segmented image of the coronary arteries after coarse segmentation of the initial vascular image.

[0071] Specifically, a pre-trained skeleton segmentation model is obtained, and the initial blood vessel image is input into the skeleton segmentation model to obtain the coronary artery skeleton image output by the skeleton segmentation model after segmenting and refining the initial blood vessel image. Optionally, based on the obtained coarse coronary artery segmentation image, a refining method, such as an optimal path algorithm, can be used to refine the coarse coronary artery segmentation image to obtain the coronary artery skeleton image. See the example below. Figure 3 , Figure 3 This is the coronary artery skeleton image corresponding to the initial blood vessel image after segmentation and refinement in this embodiment.

[0072] S120. Determine each skeleton point on the skeleton line in the coronary artery skeleton image, and determine the skeleton branch points in each skeleton point based on the degree of each skeleton point.

[0073] In this embodiment, the coronary artery skeleton image includes skeleton lines that characterize the coronary artery skeleton information. See also: Figure 3 The skeleton line can be viewed as being composed of an infinite number of skeleton points. Specifically, the degree of each skeleton point in the skeleton line is determined, and the type of each skeleton point is determined based on its degree. In this embodiment, the skeleton point types include, but are not limited to, skeleton branch points, skeleton center points, and skeleton endpoints.

[0074] It should be explained that, for any node in an undirected graph, the number of other nodes associated with that node is called the degree of the node. Based on this, the degree of a skeleton point in this embodiment can be understood as the number of points that are associated with the current skeleton point and other skeleton points on its skeleton line.

[0075] Optionally, the method for determining skeleton branch points based on the degree of skeleton points in this embodiment may include: for any skeleton point, determining each neighboring point of the current skeleton point, and determining the degree of the current skeleton point based on the skeleton points contained in the neighboring points; obtaining a preset degree threshold, and determining the skeleton branch points in each skeleton point based on the degree of the skeleton point and the preset degree threshold.

[0076] In this context, neighboring points can be understood as adjacent points of the current skeleton point in each spatial direction. Specifically, it determines whether the neighboring points of the current skeleton point coincide with other skeleton points in the current skeleton line, and determines the degree of the current skeleton point based on the determination result. For example, if the current skeleton point has 26 neighboring points, and none of the 26 neighboring points coincide with other skeleton points, then the degree of the current skeleton point is 0; conversely, if there are neighboring points that coincide with other skeleton points among the 26 neighboring points, then the degree of the current skeleton point is determined based on the number of coinciding neighboring points.

[0077] In this embodiment, the degree threshold can be understood as a threshold used to classify the type of a skeleton point. Different degree thresholds can determine different skeleton point types. Specifically, the degree threshold corresponding to a skeleton branch point is obtained, and the degree threshold and the degree of the current skeleton point are used to determine whether the current skeleton point is a skeleton branch point. Optionally, the degree of the current skeleton point is compared with the degree threshold. If the degree of the current skeleton point is within the range of the degree threshold, the current skeleton point is determined to be a skeleton branch point on the skeleton line; otherwise, the current skeleton point is determined to be a skeleton point of another type on the skeleton line.

[0078] For example, if the degree threshold of 3 is obtained for the skeleton branch point, and the degree of the current skeleton point is determined to be 3, then the current skeleton point can be determined to be a skeleton branch point based on the degree threshold; conversely, if the degree of the current skeleton point is determined to be 2, 4, or other values, then the current skeleton point can be determined to be another type of skeleton point based on the degree threshold. See also Figure 4 In the diagram, circular dots represent skeleton branch points, while rectangular and square dots represent other types of skeleton points, such as skeleton endpoints and skeleton center points. It's worth noting that the degree threshold for skeleton branch points is set to 3 because: if a skeleton point has a degree of 3, it means that the current skeleton point has three neighboring points that overlap with other skeleton points. That is, the current skeleton point has three adjacent points within the skeleton line. Based on this, it can be determined that the current skeleton point is located at a branch point within the skeleton line, thus identifying it as a skeleton branch point.

[0079] It should be noted that in practical applications, some adhesion inevitably occurs at the branching points of blood vessels, resulting in multiple branching points with more than 3 neighboring points. In this embodiment, such branching points are defined as candidate branching points. If multiple candidate branching points appear at a branching point, the candidate branching points need to be screened to obtain the final skeleton branching point, and other candidate branching points are optimized into skeleton center points to avoid bulging in the segmentation results.

[0080] Specifically, candidate branch points can be filtered based on the skeleton endpoints in the skeleton points and the skeleton connection paths between the endpoints to obtain skeleton branch points.

[0081] Optionally, the method for filtering candidate branch points to obtain skeleton branch points can be as follows: determine the skeleton endpoints and multiple candidate branch points in the skeleton points based on the degree of each skeleton point. The skeleton endpoints include the root endpoint and multiple terminal endpoints. Determine the endpoint number of each terminal endpoint, and determine the skeleton connection path between the root endpoint and any two adjacent terminal endpoints. Filter each candidate branch point based on the skeleton connection path to obtain the skeleton branch points in the candidate branch point set.

[0082] Specifically, a pre-set degree threshold of 1 is set for skeleton endpoints, and a degree threshold greater than 3 is set for candidate branch points. Further, based on the degree of the current skeleton point and the degree thresholds for each type of skeleton point, the type of the current skeleton point is determined, thereby identifying the skeleton endpoints and candidate branch points within each skeleton point. See the example below. Figure 5 , Figure 5 The multiple circular points at the bifurcation point represent multiple candidate branch points, and the square points represent skeleton endpoints. Of course, the diagram also includes other types of skeleton points, such as skeleton center points. It should be noted that the degree threshold corresponding to the skeleton endpoint is set to 1 because: if the degree of a skeleton point is 1, it means that the current skeleton point has one neighboring point that overlaps with other skeleton points. That is, the current skeleton point has one adjacent point in the skeleton line. Based on this, it can be determined that the current skeleton point is located at an endpoint in the skeleton line, thus identifying the current skeleton point as a skeleton endpoint.

[0083] In this embodiment, the root endpoint can be understood as the branch endpoint of a coronary artery branch originating from the aorta. Specifically, the root endpoint includes a left root endpoint and a right root endpoint. In practical applications, the aorta branches into two coronary artery branches on the left and right sides, respectively, in the left and right sinus regions. Therefore, the point where the left sinus region connects to the left coronary artery branch is the left root endpoint, and the point where the right sinus region connects to the right coronary artery branch is the right root endpoint. The terminal endpoint can be understood as the endpoint of the final branch of each coronary artery branch.

[0084] See examples Figure 6 , Figure 6 The thickest vessel in the middle is the aorta, which branches off from the main artery. The aorta branches into two branches: the left coronary artery and the right coronary artery. The point where the left branch connects to the aorta is the left endpoint of the root, and the point where the right branch connects to the aorta is the right endpoint of the root. (Continue reading...) Figure 6 The terminal endpoints of the branches of the coronary arteries are called terminal branch endpoints.

[0085] Optionally, the method for determining the root endpoint and at least one terminal branch endpoint among the skeleton endpoints in this embodiment may include: determining the aorta location in the coarse segmentation image of the coronary artery, and determining the root endpoint and at least one terminal branch endpoint among the skeleton endpoints based on the aorta location and the location of each skeleton endpoint.

[0086] Specifically, a pre-trained endpoint classification model can be obtained. Based on the endpoint positions of each skeletal endpoint and the arterial position of the aorta, the root endpoint and terminal branch endpoint among the skeletal endpoints can be determined using the aforementioned endpoint classification model. Optionally, the technical solution of this embodiment can also be to determine whether each skeletal endpoint is connected to the left or right sinus in the aorta. If so, it is identified as the left or right root endpoint; otherwise, it is identified as the terminal branch endpoint. This embodiment can also use other methods to identify endpoints, and the comparison is not limited.

[0087] See also the examples. Figure 5 , Figure 5 This is a schematic diagram of the left coronary artery branches. The square dot at the top right of the diagram is the root endpoint, and the other square dots of the coronary artery branches in the diagram are the terminal endpoints.

[0088] Specifically, determine the sequence number of each terminal point; see examples for further details. Figure 5 , Figure 5 The algorithm includes four terminal endpoints. Each terminal endpoint is encoded sequentially to obtain its encoding number. Further, taking terminal endpoints 3 and 4 as examples, which are adjacent terminal endpoints, the skeleton connection paths between the root left endpoint and terminal endpoint 3, and between the root left endpoint and terminal endpoint 4, are determined respectively. Based on these determined skeleton connection paths, the skeleton branch points among the candidate branch points at the branch locations are then determined.

[0089] Optionally, the method for determining the skeleton branch point among the candidate branch points at the branch in the path can be: determining each candidate branch point that is repeatedly traversed in each skeleton connected path, and determining the skeleton distance between each candidate branch point and the root endpoint; and determining the skeleton branch point among the candidate branch points based on the comparison results of each skeleton distance.

[0090] Specifically, for the skeleton connected paths corresponding to the two terminal endpoints mentioned above, candidate branch points at the branch points in the paths are determined. Further, candidate branch points that are repeatedly traversed on the two skeleton connected paths are determined, and the skeleton distance from each candidate branch point to the root endpoint is determined. Optionally, if multiple candidate branch points are repeatedly traversed, the candidate branch point corresponding to the smallest skeleton distance is determined as the skeleton branch point. See the example below. Figure 7 , Figure 7 The candidate branch points pointed to by the middle arrow are the selected skeleton branch points. Optionally, based on the above method, all candidate branch points contained in each skeleton point are filtered to obtain skeleton branch points, and other candidate branch points are optimized into skeleton center points.

[0091] S130. Based on the location of the branch points of the skeleton in the coarse segmentation image / coronary skeleton image of the coronary artery, identify the vein branch images in the initial vascular image.

[0092] In this embodiment of the invention, venous branches are identified based on the location of the branch points within the skeleton points, after determining each branch point in the skeleton. Optionally, identification can be based on the location of the branch points in the coarse segmentation image of the coronary artery, or it can be based on the location of the branch points in the coronary artery skeleton image, to obtain a venous branch image.

[0093] Optionally, based on the branch point positions of the skeleton branch points in the coarse segmentation image of the coronary artery, the vein branch image in the initial vascular image is identified, including: determining the branch point segmentation positions of the skeleton branch points in the coarse segmentation image of the coronary artery, and determining the coronary artery branch region in the coarse segmentation image of the coronary artery based on the branch point segmentation positions; obtaining a pre-trained vein branch recognition model, and performing vein branch recognition on the coronary artery branch region based on the vein branch recognition model to obtain the vein branch image.

[0094] In this context, the branch point segmentation location can be understood as the segmentation center point corresponding to the skeleton branch point in the coarse segmentation image of the coronary artery. The coronary artery branch region can be understood as the region of interest in the coarse segmentation image of the coronary artery, centered on the skeleton branch point and containing a preset number of pixels.

[0095] Specifically, the branch point locations of the skeleton branch points within the coronary artery skeleton line are determined. Based on the correspondence between the coronary artery skeleton line and the coarse segmentation image of the coronary artery, and the determined branch point locations, the segmentation positions of the skeleton branch points in the coarse segmentation image of the coronary artery are determined. Then, using this segmentation position as the center point, a region of interest of a preset size is obtained, and this region of interest is taken as the coronary artery branch region. Further, a pre-trained vein branch recognition model is obtained. The aforementioned coronary artery branch region is input into the vein branch recognition model, and vein branch recognition is performed on the coronary artery branch region based on the vein branch recognition model to obtain the vein branch image output by the model.

[0096] See examples Figure 8 , Figure 8 This represents the coronary artery branching region determined based on the branch points. Specifically, as shown in the image, the coronary artery branching region contains multiple vascular branches bounded by branch points. Each vascular branch may be a venous branch, a coronary artery branch, or both. Therefore, vein identification is required for each vascular branch within the coronary artery branching region. Based on the vein identification results, further processing of the coarse coronary artery segmentation image is performed to obtain a precise coronary artery segmentation result.

[0097] Specifically, the above Figure 8The coronary artery branch regions in the image are input into a pre-trained vein branch recognition model. The model then identifies each vessel branch, obtaining the vein branch recognition result for the current coronary artery branch region. Further, based on the recognition results of each coronary artery branch region, the vein branch image corresponding to the coarse segmentation image of the coronary artery is obtained. For example, the vein branch image is as follows: Figure 9 As shown.

[0098] Optionally, in this embodiment, the method for identifying vein branches in an initial vascular image based on the branch point position of the skeleton branch point in the coronary artery skeleton image may include: determining the skeleton position of the skeleton branch point in the coronary artery skeleton image, and determining the skeleton branch region in the coronary artery skeleton image based on the skeleton position of the branch point; obtaining a pre-trained vein branch skeleton recognition model, performing vein branch skeleton recognition on the skeleton branch region based on the vein branch skeleton recognition model, and obtaining the vein branch skeleton; obtaining the branch expansion parameters of the vein branch skeleton, and performing expansion processing on the vein branch skeleton based on the branch expansion parameters, and obtaining the vein branch image corresponding to the vein branch skeleton.

[0099] The branch point skeleton position can be understood as the location of the skeleton branch point within the coronary artery skeleton line. The skeleton branch region can be understood as the region of interest in the coronary artery skeleton image, centered on the skeleton branch point and containing a preset number of pixels.

[0100] Specifically, the branching point of the skeleton is determined within the coronary artery skeleton line. Then, using this branching point as the center point, a region of interest of a preset size is obtained, and this region of interest is used as the skeleton branching region. For example, the skeleton branching region can be... Figure 10 As shown. See also Figure 10 The skeleton branch region contains multiple skeleton branches bounded by branch points. Each skeleton branch may be a venous branch, a coronary artery branch, or both. Therefore, vein identification is required for each skeleton branch in the skeleton branch region. Based on the vein identification results, the coarse coronary artery segmentation image can be further processed to obtain accurate coronary artery segmentation results.

[0101] Furthermore, a pre-trained vein branch recognition model is obtained. The aforementioned skeleton branch region is input into the vein branch recognition model, and vein branches are identified in the skeleton branch region based on the model to obtain the vein branch recognition result for the current skeleton branch region. Further, the vein branch skeleton in the coronary artery skeleton line is obtained based on the recognition results of each skeleton branch region. For example, the identified vein branch skeleton is shown below. Figure 11 As shown.

[0102] Furthermore, the vein branch skeleton is expanded based on the branch expansion parameters to obtain a vein branch image.

[0103] In this embodiment, the branch expansion parameter can be understood as the expansion diameter parameter that the venous branch skeleton needs to be expanded. The branch expansion parameter may include the endpoint expansion parameter and the central expansion parameter of the venous branch skeleton.

[0104] It should be noted that, in this embodiment, the venous branch image obtained after expanding the venous branch skeleton based on the expansion parameters must completely cover the venous branches in the coarse coarse segmentation image of the coronary artery. Only under these conditions can vein removal processing be performed on the venous branches in the coarse coarse segmentation image of the coronary artery based on the venous branch image. Therefore, it is necessary to determine the branch expansion parameters for expanding the venous branch skeleton based on the coarse segmentation results in the coarse coarse segmentation image of the coronary artery.

[0105] Specifically, a coarse segmentation image of the coronary arteries can be obtained by performing coarse segmentation of the initial vascular image using existing segmentation techniques, or by performing segmentation processing on the initial vascular image using traditional image processing algorithms. This embodiment does not limit the segmentation method.

[0106] Furthermore, based on the determination of the venous branch skeleton, the venous branch skeleton is expanded to obtain a venous branch image. Optionally, the expansion method in this embodiment may include: for any endpoint of any venous branch skeleton, the endpoint expansion parameter is determined based on the line segment endpoint parameters of the venous branch skeleton and the vessel parameters of other vessels connected to the venous branch in the coarse segmentation image of the coronary artery. Further, based on the determination of the endpoint expansion parameters of the two line segment endpoints of the venous branch skeleton, the expansion parameters of the midpoint of the venous branch skeleton are fitted using the two endpoint expansion parameters to determine at least one center point expansion parameter of the venous branch skeleton. Then, the endpoint expansion parameters and center point expansion parameters are used as expansion targets to expand the venous branch skeleton, obtaining a venous branch image of the venous branch skeleton. Optionally, based on the above implementation method, all venous branch skeletons are traversed to obtain... Figure 9 The image shown is of a vein branch.

[0107] S140. Determine the coronary artery segmentation image based on the coarse segmentation image of the coronary artery and the venous branch image.

[0108] In this embodiment of the invention, based on the determined venous branch image and coarse segmentation image of the ...

[0109] Specifically, it can involve identifying the branch regions in the coarse segmentation image of the coronary artery that overlap with the venous branch image, and removing these branch regions from the coronary artery segmentation image. The resulting image is the coronary artery segmentation image with the venous branches removed.

[0110] For example, in obtaining such Figure 9 Based on the venous branch image shown, based on the venous branch image, Figure 2 The coarse segmentation image of the coronary artery shown is processed by removing vein branches to obtain the following result: Figure 12 The image shown is a segmented image of the coronary arteries.

[0111] The technical solution provided by this invention involves acquiring an initial vascular image, determining a coarse coarse segmentation image and a coarse skeleton image of the initial vascular image, determining each skeleton point on the skeleton line in the coarse skeleton image, and determining skeleton branch points within each skeleton point based on the degree of each skeleton point, identifying venous branch images in the initial vascular image based on the branch point positions of the skeleton branch points in the coarse coarse segmentation image / coronary skeleton image, and determining a coarse segmentation image based on the coarse coarse segmentation image and the venous branch images. This technical solution processes the vascular image to be segmented to obtain each endpoint in the blood vessel, determines candidate branch points based on each endpoint, further determines each branch point, identifies the blood vessels between each branch point to obtain venous branches, and then optimizes the coarse coarse segmentation result based on the identified venous branches to obtain the final coarse coarse segmentation result, thereby improving the accuracy of coarse coarse segmentation.

[0112] Figure 13 This is a schematic diagram of a blood vessel image segmentation device provided in an embodiment of the present invention. Figure 13 As shown, the device includes: an image determination module 210, a skeleton branch point determination module 220, a vein branch image determination module 230, and a coronary artery segmentation image determination module 240; wherein,

[0113] Image determination module 210 is used to acquire an initial blood vessel image and determine the coronary artery coarse segmentation image and coronary artery skeleton image of the initial blood vessel image, respectively;

[0114] The skeleton branch point determination module 220 is used to determine each skeleton point on the skeleton line in the coronary skeleton image, and to determine the skeleton branch point in each skeleton point based on the degree of each skeleton point.

[0115] The vein branch image determination module 230 is used to identify vein branch images in the initial blood vessel image based on the branch point positions of the skeleton branch points in the coarse segmentation image of the coarse ...

[0116] The coronary artery segmentation image determination module 240 is used to determine the coronary artery segmentation image of the initial blood vessel image based on the coarse coronary artery segmentation image and the venous branch image.

[0117] Based on the above embodiments, optionally, the image determination module 210 includes:

[0118] The coarse segmentation image acquisition unit is used to input the initial blood vessel image into a pre-trained coarse segmentation model of the coarse segmentation ...

[0119] The coronary artery skeleton image acquisition unit is used to input the initial blood vessel image into a pre-trained skeleton segmentation model to obtain the coronary artery skeleton image output by the skeleton segmentation model.

[0120] Based on the above implementation, optionally, the skeleton branch point determination module 220 includes:

[0121] The skeleton point degree determination unit is used to determine each neighboring point of the current skeleton point for any skeleton point, and determine the degree of the current skeleton point based on the skeleton points contained in the neighboring points.

[0122] The skeleton point type determination unit is used to obtain a preset degree threshold and determine the skeleton branch points in each skeleton point based on the degree of the skeleton point and the preset degree threshold.

[0123] Based on the above implementation, optionally, the skeleton branch point determination module 220 includes:

[0124] An endpoint and candidate branch point determination unit is used to determine the skeleton endpoints and multiple candidate branch points in the skeleton points based on the degree of each skeleton point. The skeleton endpoints include root endpoints and multiple terminal branch endpoints.

[0125] The skeleton branch point determination unit is used to determine the endpoint number of each of the last branch endpoints, and to determine the skeleton connection path between the root endpoint and any two adjacent last branch endpoints, and to filter each candidate branch point based on the skeleton connection path to obtain the skeleton branch points in the candidate branch point set.

[0126] Based on the above embodiments, optionally, the skeleton branch point determination unit includes:

[0127] The skeleton distance determination subunit is used to determine each candidate branch point that is repeatedly traversed in each skeleton connected path, and to determine the skeleton distance between each candidate branch point and the root endpoint respectively.

[0128] The skeleton branch point determination unit is used to determine the skeleton branch point among the candidate branch points based on the comparison results of each skeleton distance.

[0129] Based on the above embodiments, optionally, the vein branch image determination module 230 includes:

[0130] A coronary artery branch region determination unit is used to determine the branch point segmentation position of the skeleton branch point in the coarse segmentation image of the coronary artery, and to determine the coronary artery branch region in the coarse segmentation image of the coronary artery based on the branch point segmentation position.

[0131] The first vein branch image acquisition unit is used to acquire a pre-trained vein branch recognition model, and to perform vein branch recognition on the coronary artery branch region based on the vein branch recognition model to obtain a vein branch image.

[0132] Based on the above embodiments, optionally, the vein branch image determination module 230 includes:

[0133] A skeleton branch region determination unit is used to determine the skeleton branch point in the coronary artery skeleton image, and to determine the skeleton branch region in the coronary artery skeleton image based on the skeleton branch point position.

[0134] The vein branch skeleton determination unit is used to acquire a pre-trained vein branch skeleton recognition model, and to perform vein branch skeleton recognition on the skeleton branch region based on the vein branch skeleton recognition model to obtain the vein branch skeleton.

[0135] The second vein branch image acquisition unit is used to acquire the branch expansion parameters of the vein branch skeleton, and to perform expansion processing on the vein branch skeleton based on the branch expansion parameters to obtain the vein branch image corresponding to the vein branch skeleton.

[0136] The blood vessel image segmentation device provided in the embodiments of the present invention can execute the blood vessel image segmentation method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method.

[0137] Figure 14 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0138] like Figure 14As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0139] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0140] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as blood vessel image segmentation methods.

[0141] In some embodiments, the blood vessel image segmentation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the blood vessel image segmentation method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the blood vessel image segmentation method by any other suitable means (e.g., by means of firmware).

[0142] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0143] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0144] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0145] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0146] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0147] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0148] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0149] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for segmenting blood vessel images, characterized in that, include: Acquire an initial vascular image, and determine the coarse segmentation image of the coronary artery and the coronary artery skeleton image of the initial vascular image respectively; Determine each skeleton point on the skeleton line in the coronary artery skeleton image, and determine the skeleton branch points in each skeleton point based on the degree of each skeleton point; Based on the branch point positions of the skeleton branch points in the coarse segmentation image of the coronary artery / the coronary artery skeleton image, vein branch images in the initial vascular image are identified; The coronary artery segmentation image is determined based on the coarse segmentation image of the coronary artery and the venous branch image; The step of determining the skeleton branch points among the skeleton points based on the degree of each skeleton point further includes: Based on the degree of each skeleton point, determine the skeleton endpoints and multiple candidate branch points in the skeleton points, wherein the skeleton endpoints include the root endpoint and multiple terminal branch endpoints; Determine the endpoint number of each of the last branch endpoints, and determine the skeleton connection path between the root endpoint and any two adjacent last branch endpoints. Based on the skeleton connection path, filter each candidate branch point to obtain the skeleton branch points in the candidate branch point set.

2. The method according to claim 1, characterized in that, The process of determining the coarse segmentation image of the coronary arteries and the coronary artery skeleton image of the initial vascular image includes: The initial blood vessel image is input into a pre-trained coarse coarse segmentation model of the ... The initial blood vessel image is input into a pre-trained skeleton segmentation model to obtain the coronary artery skeleton image output by the skeleton segmentation model.

3. The method according to claim 1, characterized in that, The step of determining the skeleton branch points among the skeleton points based on the degree of each skeleton point includes: For any skeleton point, determine each neighboring point of the current skeleton point, and determine the degree of the current skeleton point based on the skeleton points contained in the neighboring points; Obtain a preset degree threshold, and determine the skeleton branch points in each skeleton point based on the degree of the skeleton point and the preset degree threshold.

4. The method according to claim 1, characterized in that, Filtering at least one candidate branch point based on the skeleton connectivity path includes: Determine each candidate branch point that is repeatedly traversed in each skeleton connected path, and determine the skeleton distance between each candidate branch point and the root endpoint respectively; Based on the comparison results of each skeleton distance, skeleton branch points among the candidate branch points are determined.

5. The method according to claim 1, characterized in that, The step of identifying venous branch images in the initial vascular image based on the branch point positions of the skeleton branch points in the coarse segmentation image of the coronary artery includes: The branch point of the skeleton is determined in the coarse segmentation image of the coronary artery, and the coronary artery branch region in the coarse segmentation image of the coronary artery is determined based on the branch point segmentation position; A pre-trained vein branch recognition model is obtained, and vein branches are identified in the coronary artery branch region based on the vein branch recognition model to obtain a vein branch image.

6. The method according to claim 1, characterized in that, The step of identifying venous branch images in the initial vascular image based on the branch point positions of the skeleton branch points in the coronary artery skeleton image includes: Determine the location of the branch point skeleton in the coronary artery skeleton image, and determine the skeleton branch region in the coronary artery skeleton image based on the location of the branch point skeleton. A pre-trained vein branch skeleton recognition model is obtained, and vein branch skeleton recognition is performed on the skeleton branch region based on the vein branch skeleton recognition model to obtain the vein branch skeleton; Obtain the branch expansion parameters of the vein branch skeleton, and perform expansion processing on the vein branch skeleton based on the branch expansion parameters to obtain the vein branch image corresponding to the vein branch skeleton.

7. A blood vessel image segmentation device, characterized in that, include: The image determination module is used to acquire an initial blood vessel image and determine the coarse segmentation image of the coronary artery and the coronary artery skeleton image of the initial blood vessel image, respectively. The skeleton branch point determination module is used to determine each skeleton point on the skeleton line in the coronary skeleton image, and to determine each skeleton endpoint and candidate branch point set in each skeleton point based on the degree of each skeleton point, and to determine the skeleton branch point in the candidate branch point set based on each skeleton endpoint. The vein branch image determination module is used to identify vein branch images in the initial blood vessel image based on the branch point positions of the skeleton branch points in the coarse segmentation image of the coarse ... A coronary artery segmentation image determination module is used to determine the coronary artery segmentation image of the initial blood vessel image based on the coarse coronary artery segmentation image and the venous branch image; The skeleton branch point determination module further includes: An endpoint and candidate branch point determination unit is used to determine the skeleton endpoints and multiple candidate branch points in the skeleton points based on the degree of each skeleton point. The skeleton endpoints include root endpoints and multiple terminal branch endpoints. The skeleton branch point determination unit is used to determine the endpoint number of each of the last branch endpoints, and to determine the skeleton connection path between the root endpoint and any two adjacent last branch endpoints, and to filter each candidate branch point based on the skeleton connection path to obtain the skeleton branch points in the candidate branch point set.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the blood vessel image segmentation method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the blood vessel image segmentation method according to any one of claims 1-6.

Images