Vessel image processing method, apparatus, device, medium and product

Abstract

Embodiments of the present application disclose a blood vessel image processing method, device, equipment, medium and product, wherein the method comprises: acquiring three-dimensional blood vessel image data, and performing skeletonization processing on the three-dimensional blood vessel image data to obtain a corresponding blood vessel center line; converting voxel coordinates of the blood vessel center line into physical space coordinates; creating a three-dimensional grid of the blood vessel center line based on the physical space coordinates to obtain a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data. The technical scheme of the embodiments of the present application solves the problem that the existing blood vessel center line processing may cause topological fracture of the blood vessel center line, can maintain the connectivity of the blood vessels in the blood vessel center line extraction result, and the center line processing result can be directly used for 3D printing without additional processing.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of image processing, and particularly relate to a blood vessel image processing method, device, equipment, medium and product. BACKGROUND

[0002] In the blood vessel image processing scenario, the current mainstream blood vessel centerline extraction method is mostly based on distance transformation centerline extraction method (such as Euclidean distance transformation), centerline extraction based on thinning algorithm (such as Zhang-Suen thinning algorithm) or centerline tracking based on minimum path method.

[0003] However, the above method is prone to breakpoints at complex bifurcations of blood vessels, resulting in easy breakage of the blood vessel topological structure. SUMMARY

[0004] Embodiments of the present application provide a blood vessel image processing method, device, equipment, medium and product, which can maintain the connectivity of the blood vessel centerline extraction result, and the centerline processing result can be directly used for 3D printing without additional processing.

[0005] In a first aspect, embodiments of the present application provide a blood vessel image processing method, which comprises:

[0006] obtaining three-dimensional blood vessel image data, and performing skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel centerlines;

[0007] converting voxel coordinates of the blood vessel centerlines into physical space coordinates;

[0008] creating a three-dimensional grid of the blood vessel centerlines based on the physical space coordinates to obtain a target blood vessel centerline model file corresponding to the three-dimensional blood vessel image data.

[0009] In a second aspect, embodiments of the present application provide a blood vessel image processing device, which comprises:

[0010] a centerline extraction module configured to obtain three-dimensional blood vessel image data, and perform skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel centerlines;

[0011] a coordinate conversion module configured to convert voxel coordinates of the blood vessel centerlines into physical space coordinates;

[0012] a centerline model file output module configured to create a three-dimensional grid of the blood vessel centerlines based on the physical space coordinates to obtain a target blood vessel centerline model file corresponding to the three-dimensional blood vessel image data.

[0013] In a third aspect, embodiments of the present application further provide a computer device, which comprises:

[0014] one or more processors;

[0015] a memory for storing one or more programs;

[0016] When the one or more programs are executed by the one or more processors, the one or more processors implement the blood vessel image processing method provided by any of the embodiments of the present application.

[0017] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, having stored thereon a computer program, which, when executed by a processor, implements the blood vessel image processing method provided by any of the embodiments of the present application.

[0018] In a fifth aspect, the embodiments of the present application further provide a computer program product, comprising a computer program, which, when executed by a processor, implements the blood vessel image processing method provided by any of the embodiments of the present application.

[0019] The embodiments of the above application have the following advantages or beneficial effects:

[0020] The technical scheme of the embodiments of the present application obtains three-dimensional blood vessel image data, and performs skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines; converts the voxel coordinates of the blood vessel center lines into physical space coordinates; creates a three-dimensional grid of the blood vessel center lines based on the physical space coordinates to obtain a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data. The technical scheme of the embodiments of the present application solves the problem that the existing blood vessel center line processing may cause topological fracture of the blood vessel center line, can maintain the connectivity of the blood vessels in the blood vessel center line extraction result, and the center line processing result can be directly used for 3D printing without additional processing. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a flowchart of a blood vessel image processing method provided by an embodiment of the present application;

[0022] Figure 2 is a schematic diagram of a blood vessel center line extraction process provided by an embodiment of the present application;

[0023] Figure 3 is a schematic diagram of a blood vessel image processing effect provided by an embodiment of the present application;

[0024] Figure 4 is a flowchart of another blood vessel image processing method provided by an embodiment of the present application;

[0025] Figure 5 is a structural schematic diagram of a blood vessel image processing device provided by an embodiment of the present application;

[0026] Figure 6Fig. 1 is a structural schematic diagram of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] The present application will be further described below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are intended to be merely illustrative of the present application and not in limitation thereof. It should also be noted that, for the purpose of description, only the parts related to the present application are shown in the drawings rather than all the parts.

[0028] Figure 1 Fig. 1 is a structural schematic diagram of a computer device according to an embodiment of the present application.

[0029] As shown in Fig. 1, the computer device according to the present embodiment includes a processor 10, a memory 20, a communication interface 30, and a display 40. Figure 1 The blood vessel image processing method according to the present embodiment includes the following steps:

[0030] S110, obtaining three-dimensional blood vessel image data and performing skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines.

[0031] The three-dimensional blood vessel image data can be segmented and labeled three-dimensional blood vessel image data, and the data format can be a text-based medical image format (Nearly Raw Raster Data, NRRD,.nrrd). The three-dimensional blood vessel image data can be represented by a three-dimensional array.

[0032] The process of obtaining three-dimensional blood vessel image data can be to read the corresponding blood vessel image data file from a preset image data storage path, thereby obtaining the corresponding three-dimensional blood vessel image data for subsequent blood vessel center line processing operations or other image processing operations. In addition to the three-dimensional array for representing the three-dimensional blood vessel voxel data, the blood vessel image data file can also contain meta-information for describing the array, such as the spatial origin, direction, and pixel spacing corresponding to the three-dimensional array, etc.

[0033] In the present embodiment, the skeletonization processing is performed on the three-dimensional blood vessel image data in a skeletonization manner to extract the center line of the three-dimensional blood vessel image data. The skeletonization processing is to simplify the binary three-dimensional blood vessel into a one-dimensional center line structure through iterative erosion, thereby determining the blood vessel center line. This manner can maintain the topological connectivity of the original blood vessel and avoid the generation of breakpoints in the blood vessel topological structure.

[0034] In an alternative embodiment, in order to obtain more accurate blood vessel centerline, the embodiment adopts a combination of skeletonization processing and machine learning model to determine the coordinates of the blood vessel centerline.

[0035] Specifically, the centerline skeleton in the three-dimensional blood vessel image data can be extracted based on a preset centerline skeleton extraction algorithm; then, the centerline skeleton is input into a pre-trained blood vessel centerline coordinate recognition model to obtain blood vessel centerline coordinates output by the model, and a blood vessel centerline is determined based on the blood vessel centerline coordinates.

[0036] The pre-trained blood vessel centerline coordinate recognition model is obtained by taking the coordinate-labeled blood vessel skeleton data as a model training sample and is trained to have a blood vessel centerline coordinate recognition prediction function. The blood vessel centerline coordinate recognition model can include a feature extraction layer, a feature fusion layer, and a full connection layer. The feature extraction layer includes four or more feature extraction networks, and the feature extraction network can be a dilated convolution network with different dilation rates. Dilation convolution (also known as dilated convolution) is a special convolution operation that increases the receptive field by inserting "holes" (i.e., skipping some positions) in the standard convolution kernel while keeping the parameter amount and computational amount unchanged. The dilation rate determines the size of the hole, i.e., the spacing between the convolution kernel parameters. The combination of different dilation rates can capture both local details and global context through multi-scale feature fusion, and is widely used in tasks that require both accuracy and computational efficiency. It can help to extract the features of different blood vessels and improve the accuracy of centerline extraction. In a specific example, the dilation rates of the four feature extraction networks can be 1, 2, 4, and 8, and the convolution kernel sizes can be 33, 53, 73, and 93, respectively.

[0037] The process of inputting the centerline skeleton into the pre-trained blood vessel centerline coordinate recognition model to obtain the blood vessel centerline coordinates output by the model can refer to the process shown in Figure 2 .

[0038] Specifically comprising: inputting the center line skeleton (skeletonize-3d) into a feature extraction layer (N0, N1, N2 and N3) of a blood vessel center line coordinate identification model to obtain a plurality of hierarchical feature extraction results (NxC1, NxC2, NxC3 and NxC4), wherein NxCi represents the feature dimension extracted at the layer, N is the number of skeleton points, and Ci (i is 1, 2, 3 and 4) is the number of channels extracted at the layer. Further, the feature extraction results are fused to obtain center line skeleton fusion features; the center line skeleton fusion features are input into a fully connected layer of the blood vessel center line coordinate identification model to obtain blood vessel center line coordinates output by the fully connected layer. In this process, skip connection can be used to supplement the original information, and after all the features are integrated, a unified point feature representation NxC is output, which is finally sent to the fully connected layer (Final Prediction, final prediction module, FP module) for task output to obtain the blood vessel center line coordinates.

[0039] S120, convert the voxel coordinates of the blood vessel center line into physical space coordinates.

[0040] The voxel coordinates of the blood vessel center line are image data stored in a three-dimensional array in a computer, and the position of each voxel is represented by an integer index (i, j, k) (corresponding to the row, column and layer directions of the x, y and z axes, respectively), which only represents the position of the data in the array and has no actual physical meaning. The physical space coordinates (Physical Coordinates) are coordinates in the real world (usually in millimeters), which reflect the actual position of the voxel in the patient's body and need to be calculated in combination with the parameters during image acquisition. The core formula for converting voxel coordinates (i, j, k) to physical coordinates (x, y, z) is: physical coordinates = voxel coordinates × voxel spacing + image origin.

[0041] The physical coordinates can ensure that the target blood vessel center line model file generated in the next step is consistent with the size and position of the patient's real anatomical structure, and can be used for surgical planning, 3D printing and other practical scenarios, which can overcome the deformation defect caused by linear scaling in traditional methods.

[0042] S130, create a three-dimensional grid of the blood vessel center line based on the physical space coordinates to obtain a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data.

[0043] A triangular mesh calculation component can be called to extract an isosurface based on the physical space coordinates of the blood vessel center line to obtain vertex coordinates. Three fixed point coordinates can determine a triangular patch, and finally the vertex coordinates and the corresponding triangular face index are obtained.

[0044] The target blood vessel centerline model file is a model file that can be used for 3D printing or visualization, such as an STL (Stereolithography) model file.

[0045] The technical solution of the embodiment obtains three-dimensional blood vessel image data, performs skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel centerlines, converts the voxel coordinates of the blood vessel centerlines into physical space coordinates, creates a three-dimensional grid of the blood vessel centerlines based on the physical space coordinates, and obtains a target blood vessel centerline model file corresponding to the three-dimensional blood vessel image data. The technical solution of the embodiment solves the problem of topological fracture of the blood vessel centerline in the existing blood vessel centerline processing, can maintain the connectivity of the blood vessels in the blood vessel centerline extraction result, and the centerline processing result can be directly used for 3D printing without additional processing.

[0046] In the embodiment, obtaining a target blood vessel centerline model file corresponding to three-dimensional blood vessel image data can be an end-to-end process as shown in Figure 3 , inputting three-dimensional blood vessel image data (Segmentation Annotation, SA segmentation annotation data) to a blood vessel image processing device to obtain a corresponding target blood vessel centerline model file.

[0047] Figure 4 A flowchart of a blood vessel image processing method provided by the embodiment is provided. The embodiment and the blood vessel image processing method in the above embodiment belong to the same inventive concept, and further describes the process of obtaining a blood vessel centerline. The method can be executed by a blood vessel image processing device, which can be realized by software and / or hardware and integrated into a computer device with application development function.

[0048] As shown in Figure 4 , the blood vessel image processing method of the embodiment includes the following steps:

[0049] S210, obtaining three-dimensional blood vessel image data, performing binaryzation processing on the three-dimensional blood vessel image data to obtain binaryzation three-dimensional blood vessel image data.

[0050] The three-dimensional blood vessel image data is blood vessel segmentation data, which can be represented as I(x, y, z) (three-dimensional voxel matrix). The binaryzation processing process can be represented as: In the embodiment, the segmentation threshold can be set to 0.5 to ensure accurate distinction between blood vessels and background.

[0051] S220, respectively eroding the binaryzation three-dimensional blood vessel image data on 8 subdomains with rotational symmetry in the three-dimensional image space corresponding to the binaryzation three-dimensional blood vessel image data to obtain corresponding erosion results.

[0052] In 3D space, 8 rotationally symmetric sub-domains are usually divided based on the coordinate axis direction, for example: R1: x≥0, y≥0, z≥0; R2: x≥0, y≥0, z<0; R3: x≥0, y<0, z≥0; R4: x≥0, y<0, z<0; R5: x<0, y≥0, z≥0; R6: x<0, y≥0, z<0; R7: x<0, y<0, z≥0; R8: x<0, y<0, z<0.

[0053] In each sub-domain, the 26-neighborhood connectivity detection can be represented by iterative erosion as follows: where p represents the current voxel, represents the 26-neighborhood voxel set of voxel , represents the neighborhood voxel value, and k represents the connectivity protection threshold.

[0054] The erosion of the 8 sub-domains can be calculated in parallel, improving efficiency.

[0055] S230, taking the intersection of the erosion results to obtain the centerline skeleton.

[0056] The 8 erosion results (denoted as , , , , , , , ). Each erosion result "filters" the voxels that may be retained from a specific angle. Taking the intersection means a logical intersection operation, that is, only the voxels that exist simultaneously in multiple erosion results are retained. This skeleton extraction process can reduce the topological error rate and keep the skeleton connectivity rate >90.2%.

[0057] S240, inputting the centerline skeleton into a pre-trained blood vessel centerline coordinate recognition model to obtain the blood vessel centerline coordinates output by the model, and determining the blood vessel centerline based on the blood vessel centerline coordinates.

[0058] S250, converting the voxel coordinates of the blood vessel centerline into physical space coordinates.

[0059] The mapping relationship between the voxel coordinates and the physical space coordinates can be represented as follows:

[0060] wherein represents a spatial direction matrix; = , represents the physical space origin coordinates. represents the coordinates of the blood vessel centerline, represent physical space coordinates.

[0061] S260, create a three-dimensional grid of the vessel centerline based on the physical space coordinates, to obtain a target vessel centerline model file corresponding to the three-dimensional vessel image data.

[0062] The triangular mesh construction can be represented as . Wherein F represents the number of triangular facets, represent physical space vertex coordinates.

[0063] The technical scheme of the embodiment, by acquiring three-dimensional vessel image data, performing binaryzation processing on the three-dimensional vessel image data to obtain binaryzation three-dimensional vessel image data; respectively performing corrosion on the binaryzation three-dimensional vessel image data on the 8 sub-domains with rotational symmetry corresponding to the three-dimensional image space of the binaryzation three-dimensional vessel image data, to obtain the corresponding corrosion results; taking the intersection of the corrosion results to obtain the centerline skeleton; inputting the centerline skeleton into the pre-trained vessel centerline coordinate recognition model to obtain the vessel centerline coordinates output by the model, and determining the vessel centerline based on the vessel centerline coordinates; converting the voxel coordinates of the vessel centerline into physical space coordinates; creating a three-dimensional grid of the vessel centerline based on the physical space coordinates, to obtain a target vessel centerline model file corresponding to the three-dimensional vessel image data. The technical scheme of the embodiment solves the problem that the existing vessel centerline processing may cause topological fracture of the vessel centerline, can maintain the connectivity of the vessel in the vessel centerline extraction result, and the centerline processing result can be directly used for 3D printing without additional processing.

[0064] The scheme of the above embodiment can generate a model that can be directly used for 3D printing or visualization, efficiently process medical image data, and be applied to scenes such as computer-aided diagnosis of vascular diseases, three-dimensional model reconstruction of surgical planning, and generation of visualization teaching materials in medical education.

[0065] Figure 5 A structural schematic diagram of a blood vessel image processing device provided by the embodiment of the present application, the embodiment can be applied to the scene of processing blood vessel images, especially the case of image processing for blood vessel centerlines. The blood vessel image processing device can be realized by software and / or hardware, and integrated in a computer terminal device with application development function.

[0066] As shown in Figure 5 , the blood vessel image processing device includes a centerline extraction module 310, a coordinate conversion module 320, and a centerline model file output module 330.

[0067] The center line extraction module 310 is configured to obtain three-dimensional blood vessel image data, and perform skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines.

[0068] The technical scheme of the embodiment obtains three-dimensional blood vessel image data, performs skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines, converts voxel coordinates of the blood vessel center lines into physical space coordinates, creates a three-dimensional grid of the blood vessel center lines based on the physical space coordinates, and obtains a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data. The technical scheme of the embodiment can solve the problem that the existing blood vessel center line processing may cause topological fracture of the blood vessel center line, can maintain the connectivity of the blood vessel in the blood vessel center line extraction result, and can directly use the center line processing result for 3D printing without additional processing.

[0069] In an optional implementation, the center line extraction module 310 is specifically configured to:

[0070] extract a center line skeleton in the three-dimensional blood vessel image data based on a preset center line skeleton extraction algorithm;

[0071] input the center line skeleton into a pre-trained blood vessel center line coordinate recognition model to obtain blood vessel center line coordinates output by the model, and determine the blood vessel center line based on the blood vessel center line coordinates.

[0072] In an optional implementation, the center line extraction module 310 can also be configured to:

[0073] input the center line skeleton into a feature extraction layer of the blood vessel center line coordinate recognition model to obtain a plurality of hierarchical feature extraction results;

[0074] perform feature fusion on the feature extraction results to obtain center line skeleton fusion features;

[0075] input the center line skeleton fusion features into a full connection layer of the blood vessel center line coordinate recognition model to obtain blood vessel center line coordinates output by the full connection layer.

[0076] In an optional implementation, the feature extraction layer includes four or more feature extraction networks, and the feature extraction network is a dilated convolution network with different dilation rates.

[0077] In an optional implementation, the center line extraction module 310 can also be configured to:

[0078] The three-dimensional vascular image data is binarized to obtain binarized three-dimensional vascular image data;

[0079] The binarized three-dimensional blood vessel image data was eroded on eight rotationally symmetric subdomains of the three-dimensional image space corresponding to the binarized three-dimensional blood vessel image data, and the corresponding erosion results were obtained.

[0080] The centerline skeleton is obtained by taking the intersection of the corrosion results.

[0081] In one alternative implementation, the target blood vessel centerline model file is a model file that can be used for 3D printing or visualization.

[0082] The vascular image processing apparatus provided in the embodiments of the present invention can execute the vascular image processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0083] Figure 6 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Figure 6 A block diagram of an exemplary computer device 12 suitable for implementing embodiments of the present invention is shown. Figure 6 The computer device 12 shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of the present invention. The computer device 12 can be any terminal device with computing capabilities, such as intelligent controllers and servers, mobile phones, and other terminal devices.

[0084] like Figure 6 As shown, the computer device 12 is represented in the form of a general-purpose computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, memory 28, and a bus 18 connecting different system components (including memory 28 and processing unit 16).

[0085] Bus 18 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0086] Computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0087] Memory 28 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache 32. Computer device 12 can further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 can be provided for reading from and writing to a non-removable, non-volatile magnetic media (not shown and typically called a "hard drive"). Figure 6 Although not shown, computer device 12 can employ other, different, or Figure 6 additional components, such as one or more additional data storage devices (removable and / or non-removable, volatile and / or non-volatile). In this regard, storage system 34 as well as the storage of the removable media interface can be considered computer system storage media. Computer system storage media can also include, without limitation, volatile memories, non-volatile memories, magnetic media, optical media, solid state media, and any other medium that can be used to store desired program code means in the form of computer readable program instructions, data structures, program modules, and / or computer program

[0088] Program / utility 40, having a set (at least one) of program modules 42, can be stored in, for example, memory 28 by way of example, and not limitation, as well as an operating system, one or more application programs, other program modules, and program data, and can include an implementation of a network environment. Each of the operating systems, one or more application programs, other program modules, and program data or some combination thereof, can include an implementation of a network environment.

[0089] Computer device 12 can also communicate with one or more external devices 14 such as a keyboard or a pointing device, displays 24, etc.; one or more devices that enable a user to interact with computer device 12; and / or one or more devices that enable computer device 12 to communicate with one or more other computing devices. Such communication can be via input / output (I / O) interfaces 22. Similarly, computer device 12 can communicate with one or more networks such as a local area network (LAN), a general wide area network (WAN), and / or a public network such as the Internet, via network adapter 20. As indicated, network adapter 20 can be communicatively coupled to bus 18 via the I / O interfaces 22. It will be appreciated that other buses and interconnections that are well known will have been used in computer device 12, but are not specifically shown, so as not to obscure the disclosure. Figure 6 Other hardware and / or software modules that can be used in the computer device 12 can also be employed in conjunction with computer device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.

[0090] The processing unit 16 executes various function applications and data processing by running programs stored in the memory 28, such as implementing the blood vessel image processing method provided in the embodiments of the present application, which comprises:

[0091] Obtaining three-dimensional blood vessel image data, and performing skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines;

[0092] Converting the voxel coordinates of the blood vessel center lines into physical space coordinates;

[0093] Creating a three-dimensional grid of the blood vessel center lines based on the physical space coordinates to obtain a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data.

[0094] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the blood vessel image processing method provided in any of the embodiments of the present application, which comprises:

[0095] Obtaining three-dimensional blood vessel image data, and performing skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines;

[0096] Converting the voxel coordinates of the blood vessel center lines into physical space coordinates;

[0097] Creating a three-dimensional grid of the blood vessel center lines based on the physical space coordinates to obtain a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data.

[0098] The computer storage medium of the embodiments of the present application can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples (non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus.

[0099] A computer readable signal medium can include a propagated data signal with computer executable progra code embodied therein. Forwarded data signals accomplished by any of a variety of means, including but not limited to electronic, optical, electromagnetic, infrared, acoustic or any combination thereof, etc. Combinations of any of the above should also be included within the scope of computer readable signal media.

[0100] Program code embodied on a computer readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire line, optical fiber cable, RF, etc., or any combination thereof.

[0101] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0102] The embodiments of the present application also provide a computer program product, comprising a computer program which, when executed by a processor, implements the blood vessel image processing method according to any of the embodiments of the present application.

[0103] Computer program products can be written in any of various programming languages, or combinations thereof, include an object oriented programming language such as Java, Smalltalk, Python, C++, and conventional procedural programming languages such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0104] Those skilled in the art will appreciate that the modules or steps of the application described above can be implemented in a general purpose computer, and can be centralized in a single computer or distributed over a network of multiple computers, and optionally, they can be implemented in program code executable by a computer, and thus can be stored in a storage device and executed by a computer, or they can be made into individual integrated circuit modules, or a plurality of modules or steps can be made into a single integrated circuit module. Thus, the application is not limited to any particular combination of hardware and software.

[0105] Note that the above are only preferred embodiments of the application and the principles of the technology applied. Those skilled in the art will understand that the application is not limited to the specific embodiments described herein, and that various obvious changes, reconfigurations and substitutions can be made by those skilled in the art without departing from the scope of the application. Therefore, although the application has been described in detail through the above embodiments, the application is not limited to the above embodiments, and can include more other equivalent embodiments without departing from the concept of the application, and the scope of the application is determined by the scope of the appended claims.

Claims

1. A blood vessel image processing method characterized by comprising: The method comprises the following steps: acquiring three-dimensional blood vessel image data, and performing skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines; converting voxel coordinates of the blood vessel center lines into physical space coordinates; creating a three-dimensional grid of the blood vessel center lines based on the physical space coordinates to obtain a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data; wherein the skeletonization processing on the three-dimensional blood vessel image data to obtain the corresponding blood vessel center lines comprises: extracting a center line skeleton in the three-dimensional blood vessel image data based on a preset center line skeleton extraction algorithm; inputting the center line skeleton into a pre-trained blood vessel center line coordinate recognition model to obtain blood vessel center line coordinates output by the model, and determining the blood vessel center lines based on the blood vessel center line coordinates; the inputting of the center line skeleton into the pre-trained blood vessel center line coordinate recognition model to obtain the blood vessel center line coordinates output by the model comprises: inputting the center line skeleton into a feature extraction layer in the blood vessel center line coordinate recognition model to obtain a plurality of hierarchical feature extraction results; performing feature fusion on the feature extraction results to obtain center line skeleton fusion features; and inputting the center line skeleton fusion features into a fully connected layer of the blood vessel center line coordinate recognition model to obtain blood vessel center line coordinates output by the fully connected layer.

2. The method of claim 1, wherein, The feature extraction layer comprises four or more feature extraction networks, and the feature extraction networks are dilated convolution networks with different dilation rates.

3. The method of claim 1, wherein, The extraction of the center line skeleton in the three-dimensional blood vessel image data based on the preset center line skeleton extraction algorithm comprises: performing binaryzation processing on the three-dimensional blood vessel image data to obtain binaryzation three-dimensional blood vessel image data; performing erosion on the binaryzation three-dimensional blood vessel image data respectively on eight sub-domains with rotational symmetry in a three-dimensional image space corresponding to the binaryzation three-dimensional blood vessel image data to obtain corresponding erosion results; taking an intersection of the erosion results to obtain the center line skeleton.

4. The method according to any one of claims 1 to 3, characterized in that, The target blood vessel center line model file is a model file that can be used for 3D printing or visualization.

5. A blood vessel image processing apparatus characterized by comprising: The method comprises the following steps: a center line extraction module for acquiring three-dimensional blood vessel image data, and performing skeletonization processing on the three-dimensional blood vessel image data to obtain corresponding blood vessel center lines; a coordinate conversion module for converting voxel coordinates of the blood vessel center lines into physical space coordinates; a center line model file output module for creating a three-dimensional grid of the blood vessel center lines based on the physical space coordinates to obtain a target blood vessel center line model file corresponding to the three-dimensional blood vessel image data; wherein the center line extraction module is specifically configured to extract a center line skeleton in the three-dimensional blood vessel image data based on a preset center line skeleton extraction algorithm; input the center line skeleton into a pre-trained blood vessel center line coordinate recognition model to obtain blood vessel center line coordinates output by the model, and determine the blood vessel center lines based on the blood vessel center line coordinates; The inputting the center line skeleton into the pre-trained blood vessel center line coordinate identification model to obtain the blood vessel center line coordinates output by the model comprises: inputting the center line skeleton into a feature extraction layer of the blood vessel center line coordinate identification model to obtain a plurality of levels of feature extraction results; performing feature fusion on the feature extraction results to obtain center line skeleton fusion features; and inputting the center line skeleton fusion features into a full connection layer of the blood vessel center line coordinate identification model to obtain the blood vessel center line coordinates output by the full connection layer.

6. A computer device, comprising: The computer device comprises: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the blood vessel image processing method as claimed in any one of claims 1-4.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the blood vessel image processing method as claimed in any one of claims 1-4.

8. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the blood vessel image processing method as claimed in any one of claims 1-4. The computer program is executed by the processor to implement the blood vessel image processing method as claimed in any one of claims 1-4.

Images