Medical image segmentation method and device, computer equipment and storage medium
Through multiple segmentation processing and merging processing, the problem of undersegmentation in cerebrovascular medical image segmentation is solved, segmentation accuracy and integrity are improved, and the accurate segmentation of cerebrovascular images is ensured.
Patent Information
- Application Number
- CN202311050630.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-18
- Publication Date
- 2025-02-21
AI Technical Summary
In the prior art, the medical image segmentation of cerebrovascular vessels has a problem of undersegmentation, especially when cerebrovascular vessels are rich and grayscale value transformation ranges are wide, the generalization performance of deep learning methods is poor, resulting in a deviation in segmentation results.
Through multiple segmentation processing, the first vascular segmentation result, the second vascular segmentation result and the third vascular segmentation result of the vascular transition area of the image to be segmented are determined, and these results are combined to obtain the final target vascular segmentation result. Specific steps include expansion processing, mask image generation, classification processing and contour update, etc.
It effectively avoids the problem of undersegment around cerebrovascular, improves the accuracy and integrity of cerebrovascular segmentation, and ensures the accurate segmentation of cerebrovascular images.
Smart Images
Figure CN119494808A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of medical image processing, and in particular to a medical image segmentation method, apparatus, computer equipment and storage medium. Background Art
[0002] In the field of neurosurgery, accurate segmentation of cerebral blood vessels from MRA (Magnetic Resonance Angiography) or CTA (Computed Tomography Angiography) imaging can assist in the diagnosis of neurosurgical diseases and provide accurate information for the formulation of neurosurgical surgery plans. Since cerebral blood vessels are relatively close to surrounding tissues, the phase difference dynamic range is small, and there is mutual entanglement and overlap between cerebral blood vessels, resulting in partial coverage of blood vessels, as well as the presence of lesions and tiny blood vessels in the brain tissue structure, it is difficult to accurately segment cerebral blood vessels.
[0003] In the prior art, deep learning methods are usually used to segment cerebral blood vessels. However, due to certain differences in the generalization performance of deep learning, especially when faced with abundant cerebral blood vessels and a wide range of vascular grayscale value changes, and the original gold standard data annotation accuracy in the deep learning method may have certain differences, resulting in deviations in the cerebral blood vessel segmentation results based on deep learning, and it is easy to have under-segmentation of a single slice around the blood vessel, which is manifested in the MPR (MultiPlanar Reconstruction) image as a large area of under-segmentation of the blood vessels on a certain section.
[0004] Therefore, there is an under-segmentation problem in the current medical image segmentation technology for cerebral blood vessels. Summary of the invention
[0005] Based on this, it is necessary to provide a medical image segmentation method, apparatus, computer equipment, computer-readable storage medium and computer program product that can avoid under-segmentation around cerebral blood vessels in order to address the above technical problems.
[0006] In a first aspect, the present application provides a medical image segmentation method. The method comprises:
[0007] Determine a first blood vessel segmentation result of the image to be segmented;
[0008] Determining a second blood vessel segmentation result of a target portion in the image to be segmented according to the first blood vessel segmentation result;
[0009] determining a blood vessel transition region of the target site according to the second blood vessel segmentation result, and determining a third blood vessel segmentation result of the blood vessel transition region;
[0010] The first blood vessel segmentation result and the third blood vessel segmentation result are combined to obtain a target blood vessel segmentation result of the image to be segmented.
[0011] In one embodiment, determining the vascular transition region of the target site according to the second vascular segmentation result includes:
[0012] performing expansion processing on the second blood vessel segmentation result to obtain an expanded second blood vessel segmentation result;
[0013] The blood vessel transition region of the target site is determined according to the dilated second blood vessel segmentation result.
[0014] In one embodiment, determining the blood vessel transition region of the target site according to the segmentation result of the expanded second blood vessel includes:
[0015] obtaining a first mask image of the blood vessel transition region according to a difference between the dilated second blood vessel segmentation result and the second blood vessel segmentation result;
[0016] The blood vessel transition region of the target part is determined from the image to be segmented according to the first mask image.
[0017] In one embodiment, determining a second blood vessel segmentation result of a target part in the image to be segmented according to the first blood vessel segmentation result includes:
[0018] Performing segmentation processing on the target part in the image to be segmented to obtain a second mask image of the target part;
[0019] The second blood vessel segmentation result of the target part is determined from the first blood vessel segmentation result according to the second mask image.
[0020] In one embodiment, determining the blood vessel transition region of the target site according to the segmentation result of the expanded second blood vessel includes:
[0021] Determining a target part image from the image to be segmented according to the second mask image;
[0022] removing the second blood vessel segmentation result from the target part image to obtain a target part image with blood vessels removed;
[0023] The blood vessel transition region of the target part is obtained according to the intersection of the expanded second blood vessel segmentation result and the blood vessel removed target part image.
[0024] In one embodiment, determining the third blood vessel segmentation result of the blood vessel transition region includes:
[0025] performing classification processing on the vascular transition region to obtain a classification result of the vascular transition region;
[0026] The third blood vessel segmentation result of the blood vessel transition area is obtained according to the classification result that meets the preset result.
[0027] In one embodiment, obtaining the third blood vessel segmentation result of the blood vessel transition region according to the classification result that meets the preset result includes:
[0028] Obtaining an initial blood vessel segmentation result of the blood vessel transition region according to the classification result that meets the preset result;
[0029] The initial blood vessel segmentation result is processed by contour updating to obtain the third blood vessel segmentation result.
[0030] In a second aspect, the present application also provides a medical image segmentation device. The device comprises:
[0031] A first segmentation module, used to determine a first blood vessel segmentation result of the image to be segmented;
[0032] A second segmentation module, used to determine a second blood vessel segmentation result of a target part in the image to be segmented according to the first blood vessel segmentation result;
[0033] a third segmentation module, configured to determine a vascular transition region of the target site according to the second vascular segmentation result, and determine a third vascular segmentation result of the vascular transition region;
[0034] The merging processing module is used to merge the first blood vessel segmentation result and the third blood vessel segmentation result to obtain a target blood vessel segmentation result of the image to be segmented.
[0035] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:
[0036] Determine a first blood vessel segmentation result of the image to be segmented;
[0037] Determining a second blood vessel segmentation result of a target portion in the image to be segmented according to the first blood vessel segmentation result;
[0038] determining a blood vessel transition region of the target site according to the second blood vessel segmentation result, and determining a third blood vessel segmentation result of the blood vessel transition region;
[0039] The first blood vessel segmentation result and the third blood vessel segmentation result are combined to obtain a target blood vessel segmentation result of the image to be segmented.
[0040] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:
[0041] Determine a first blood vessel segmentation result of the image to be segmented;
[0042] Determining a second blood vessel segmentation result of a target portion in the image to be segmented according to the first blood vessel segmentation result;
[0043] determining a blood vessel transition region of the target site according to the second blood vessel segmentation result, and determining a third blood vessel segmentation result of the blood vessel transition region;
[0044] The first blood vessel segmentation result and the third blood vessel segmentation result are combined to obtain a target blood vessel segmentation result of the image to be segmented.
[0045] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:
[0046] Determine a first blood vessel segmentation result of the image to be segmented;
[0047] Determining a second blood vessel segmentation result of a target portion in the image to be segmented according to the first blood vessel segmentation result;
[0048] determining a blood vessel transition region of the target site according to the second blood vessel segmentation result, and determining a third blood vessel segmentation result of the blood vessel transition region;
[0049] The first blood vessel segmentation result and the third blood vessel segmentation result are combined to obtain a target blood vessel segmentation result of the image to be segmented.
[0050] The above-mentioned medical image segmentation method, apparatus, computer equipment, storage medium and computer program product determine a first blood vessel segmentation result of the image to be segmented, determine a second blood vessel segmentation result of the target part in the image to be segmented according to the first blood vessel segmentation result, determine the blood vessel transition area of the target part according to the second blood vessel segmentation result, and determine a third blood vessel segmentation result of the blood vessel transition area, merge the first blood vessel segmentation result and the third blood vessel segmentation result to obtain the target blood vessel segmentation result of the image to be segmented; based on the coarse segmentation result of the blood vessels in the cerebrovascular image, the blood vessels and the blood vessel transition area of the brain tissue area in the cerebrovascular image can be determined, and then the blood vessel transition area can be finely segmented, and the coarse segmentation result and the fine segmentation result are merged to obtain the final blood vessel segmentation result, which effectively solves the under-segmentation problem existing in blood vessel segmentation. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] Figure 1 is a schematic flow chart of a medical image segmentation method in one embodiment;
[0052] Figure 2 A schematic diagram of determining a blood vessel transition region in one embodiment;
[0053] Figure 3 A schematic diagram of determining a blood vessel transition region in another embodiment;
[0054] Figure 4 is a schematic flow chart of a cerebral blood vessel segmentation method in one embodiment;
[0055] Figure 5 is a schematic diagram of a rough blood vessel segmentation result of a cerebral blood vessel image in one embodiment;
[0056] Figure 6 is a schematic diagram of a mask of a brain tissue region in one embodiment;
[0057] Figure 7 is a schematic diagram of a rough segmentation result of blood vessels in a brain tissue region in one embodiment;
[0058] Figure 8 is a schematic diagram of a flow chart of a blood vessel transition region fine segmentation process in one embodiment;
[0059] Fig. 9 is a schematic diagram of a rough segmentation result of blood vessels in a brain tissue region in another embodiment;
[0060] Fig.10 is a schematic diagram of a vascular transition region of brain tissue in one embodiment;
[0061] Fig.11 is a schematic diagram of an initial blood vessel segmentation result of a blood vessel transition region in one embodiment;
[0062] Fig.12 is a schematic diagram of a blood vessel precise segmentation result in a blood vessel transition region in one embodiment;
[0063] Fig.13 is a schematic diagram of combining a rough blood vessel segmentation result of a cerebral blood vessel image with a fine blood vessel segmentation result of a blood vessel transition region in one embodiment;
[0064] Fig.14 is a schematic diagram of cerebral blood vessels finally segmented in one embodiment;
[0065] Fig.15 is a structural block diagram of a medical image segmentation device in one embodiment;
[0066] Fig.16 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION
[0067] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0068] In one embodiment, Figure 1 As shown, a medical image segmentation method is provided. This embodiment uses the method applied to a terminal as an example. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server.
[0069] In this embodiment, the method includes the following steps:
[0070] Step S110, determining a first blood vessel segmentation result of the image to be segmented.
[0071] The image to be segmented may be a cerebral blood vessel image.
[0072] The first blood vessel segmentation result may be a rough blood vessel segmentation result of the cerebral blood vessel image.
[0073] In a specific implementation, the image to be segmented can be input into the terminal, and the terminal segments the blood vessels in the image to be segmented to obtain a first blood vessel segmentation result. The segmentation method of the image to be segmented includes but is not limited to various coarse segmentation methods such as traditional image segmentation methods and image segmentation methods based on deep learning.
[0074] In practical applications, the cerebrovascular image may include brain tissue blood vessels and peripheral blood vessels of the brain tissue (for example, scalp blood vessels and external arteries of the brain tissue). The cerebrovascular image is input into the terminal, and the terminal performs coarse segmentation on the brain tissue blood vessels and peripheral blood vessels of the brain tissue in the cerebrovascular image to obtain a coarse blood vessel segmentation result of the cerebrovascular image, i.e., a first blood vessel segmentation result.
[0075] Step S120: determining a second blood vessel segmentation result of a target part in the image to be segmented according to the first blood vessel segmentation result.
[0076] The target site may be brain tissue.
[0077] The second blood vessel segmentation result may be a rough blood vessel segmentation result of the brain tissue region.
[0078] In a specific implementation, the terminal may segment the target part in the image to be segmented to obtain a mask of the target part, and determine a second blood vessel segmentation result of the target part from the first blood vessel segmentation result according to the mask of the target part.
[0079] In practical applications, the terminal can segment the brain tissue in the cerebrovascular image to obtain a mask of the brain tissue area, and perform an AND operation on the mask of the brain tissue area and the rough segmentation result of the blood vessels in the cerebrovascular image to obtain a rough segmentation result of the blood vessels in the brain tissue area, that is, a second blood vessel segmentation result.
[0080] Step S130, determining a vascular transition region of the target part according to the second vascular segmentation result, and determining a third vascular segmentation result of the vascular transition region.
[0081] The vascular transition region may be a region adjacent to a blood vessel and a non-blood vessel in brain tissue.
[0082] The third blood vessel segmentation result may be a blood vessel precise segmentation result of the blood vessel transition region.
[0083] In a specific implementation, the terminal can determine the vascular transition region of the target part according to the second vascular segmentation result of the target part, segment the blood vessels in the vascular transition region, and obtain a third vascular segmentation result of the vascular transition region of the target part. The segmentation method of the vascular transition region includes but is not limited to various fine segmentation methods.
[0084] In practical applications, the terminal can determine the vascular transition area of the brain tissue in the cerebrovascular image by performing morphological dilation processing on the rough segmentation result of the blood vessels in the brain tissue area, and obtain the fine segmentation result of the vascular transition area by classifying the vascular transition area, for example, using a statistical model classification algorithm, i.e., a third vascular segmentation result, wherein the statistical model classification algorithm can be a Gaussian mixture model classification algorithm or a Gaussian-Rayleigh mixture model classification algorithm.
[0085] Step S140 , merging the first blood vessel segmentation result and the third blood vessel segmentation result to obtain a target blood vessel segmentation result of the image to be segmented.
[0086] The target blood vessel segmentation result may be the final blood vessel segmentation result of the cerebral blood vessel image.
[0087] In a specific implementation, the terminal may combine the first blood vessel segmentation result of the image to be segmented with the third blood vessel segmentation result of the blood vessel transition region of the target part in the image to be segmented to obtain the target blood vessel segmentation result.
[0088] In practical applications, the terminal can combine the rough segmentation result of the blood vessels in the cerebral vascular image with the fine segmentation result of the blood vessels in the transition area of the brain tissue blood vessels to obtain a refined target blood vessel segmentation result.
[0089] The above-mentioned medical image segmentation method determines a first blood vessel segmentation result of the image to be segmented, determines a second blood vessel segmentation result of the target part in the image to be segmented according to the first blood vessel segmentation result, determines the blood vessel transition area of the target part according to the second blood vessel segmentation result, and determines a third blood vessel segmentation result of the blood vessel transition area, and merges the first blood vessel segmentation result and the third blood vessel segmentation result to obtain the target blood vessel segmentation result of the image to be segmented; based on the rough segmentation result of the blood vessels in the cerebrovascular image, the blood vessels and the blood vessel transition area of the brain tissue area in the cerebrovascular image can be determined, and then the blood vessel transition area is finely segmented, and the rough segmentation result is merged with the fine segmentation result to obtain the final blood vessel segmentation result, which effectively solves the under-segmentation problem existing in blood vessel segmentation.
[0090] In one embodiment, the above step S130 may specifically include: dilating the second blood vessel segmentation result to obtain an expanded second blood vessel segmentation result; and determining a blood vessel transition region of the target site according to the expanded second blood vessel segmentation result.
[0091] In a specific implementation, the terminal can perform morphological dilation on the second blood vessel segmentation result of the target part in the image to be segmented to obtain the expanded second blood vessel segmentation result, and determine the blood vessel transition area of the target part in the image to be segmented based on the difference between the expanded second blood vessel segmentation result and the second segmentation result.
[0092] In this embodiment, the second blood vessel segmentation result is expanded to obtain an expanded second blood vessel segmentation result; the blood vessel transition area of the target part is determined according to the expanded second blood vessel segmentation result, and the blood vessel transition area of the brain tissue in the cerebrovascular image can be determined, so as to perform precise segmentation on the blood vessel transition area and avoid under-segmentation problem in cerebrovascular segmentation.
[0093] In one embodiment, the step of determining the vascular transition region of the target part according to the expanded second blood vessel segmentation result may specifically include: obtaining a first mask image of the vascular transition region according to a difference between the expanded second blood vessel segmentation result and the second blood vessel segmentation result; and determining the vascular transition region of the target part from the image to be segmented according to the first mask image.
[0094] The first mask image may be a mask corresponding to a vascular transition region of the brain tissue.
[0095] In a specific implementation, the terminal can obtain the difference between the second blood vessel segmentation result after expansion and the second blood vessel segmentation result, obtain an image of the same size as the image to be segmented, set the pixel value corresponding to the difference to 1, set other pixel values to 0, obtain a first mask image, perform an AND operation on the first mask image and the image to be segmented, and obtain the blood vessel transition area of the target part in the image to be segmented.
[0096] Figure 2 A schematic diagram of the determination of the vascular transition zone is provided. Figure 2 The rough segmentation result of the blood vessels in the brain tissue area in the cerebrovascular image can be morphologically expanded to obtain the expanded rough segmentation result of the blood vessels, and the difference set operation is performed between the expanded rough segmentation result of the blood vessels and the rough segmentation result of the blood vessels in the brain tissue area to obtain the mask of the blood vessel transition area, and the mask is ANDed with the cerebrovascular image to obtain the blood vessel transition area of the brain tissue area in the cerebrovascular image.
[0097] In this embodiment, a first mask image of the vascular transition region is obtained based on the difference between the expanded second blood vessel segmentation result and the second blood vessel segmentation result; based on the first mask image, the vascular transition region of the target part is determined from the image to be segmented, and the vascular transition region of the brain tissue in the cerebral blood vessel image can be determined, so as to perform precise segmentation on the vascular transition region and avoid under-segmentation problem in cerebral blood vessel segmentation.
[0098] In one embodiment, the above step S120 may specifically include: performing segmentation processing on the target part in the image to be segmented to obtain a second mask image of the target part; and determining a second blood vessel segmentation result of the target part from the first blood vessel segmentation result according to the second mask image.
[0099] The second mask image may be a mask corresponding to the brain tissue region.
[0100] In a specific implementation, in the image to be segmented, the terminal can segment the target part, set the pixel value of the segmented target part to 1, set other pixel values to 0, obtain a second mask image, and perform an AND operation on the second mask image and the first blood vessel segmentation result to obtain a second blood vessel segmentation result of the target part.
[0101] In practical applications, after determining the rough segmentation result of the blood vessels in the cerebrovascular image, the terminal can segment the brain tissue in the cerebrovascular image to obtain a mask of the brain tissue area, and perform an AND operation on the mask of the brain tissue area and the rough segmentation result of the blood vessels in the cerebrovascular image to obtain the rough segmentation result of the blood vessels in the brain tissue area, that is, the second blood vessel segmentation result.
[0102] In this embodiment, a second mask image of the target part is obtained by segmenting the target part in the image to be segmented; and a second blood vessel segmentation result of the target part is determined from the first blood vessel segmentation result according to the second mask image. After the first blood vessel segmentation result is determined, the second blood vessel segmentation result can be quickly determined, thereby improving the efficiency of image segmentation.
[0103] In one embodiment, the step of determining the vascular transition region of the target part according to the expanded second blood vessel segmentation result may specifically include: determining the target part image from the image to be segmented according to the second mask image; removing the second blood vessel segmentation result from the target part image to obtain the target part image with the blood vessels removed; and obtaining the vascular transition region of the target part according to the intersection of the expanded second blood vessel segmentation result and the target part image with the blood vessels removed.
[0104] The target part image may be a brain tissue image.
[0105] In a specific implementation, the terminal can perform an AND operation on the second mask image and the image to be segmented to obtain a target part image of the target part in the image to be segmented, remove the second blood vessel segmentation result from the target part image to obtain a target part image with blood vessels removed, perform an intersection operation on the expanded second blood vessel segmentation result and the target part image with blood vessels removed to obtain a blood vessel transition area of the target part.
[0106] Figure 3 Another schematic diagram of the determination of the vascular transition zone is provided. Figure 3 The brain tissue area mask and the cerebral vascular image can be ANDed to obtain the brain tissue image, and the rough segmentation result of the blood vessels in the brain tissue area can be removed from the brain tissue image to obtain the brain tissue image with the blood vessels removed. The expanded rough segmentation result of the blood vessels can be intersected with the brain tissue image with the blood vessels removed to obtain the vascular transition area of the brain tissue.
[0107] In this embodiment, a target part image is determined from the image to be segmented according to the second mask image; the second blood vessel segmentation result is removed from the target part image to obtain the target part image with blood vessels removed; and the blood vessel transition region of the target part is obtained according to the intersection of the expanded second blood vessel segmentation result and the target part image with blood vessels removed, so that the blood vessel transition region of the brain tissue in the cerebrovascular image can be determined, so as to perform precise segmentation on the blood vessel transition region and avoid under-segmentation problem in cerebrovascular segmentation.
[0108] In one embodiment, the above step S130 may specifically include: classifying the vascular transition region to obtain a classification result of the vascular transition region; and obtaining a third vascular segmentation result of the vascular transition region according to the classification result that meets a preset result.
[0109] The classification result may be a result of classifying each pixel in the blood vessel transition region as a blood vessel or a non-blood vessel.
[0110] The preset result may be a classification result that the designated pixel is a blood vessel.
[0111] In a specific implementation, each pixel in the vascular transition region can be classified to obtain a classification result of each pixel being a blood vessel or a non-blood vessel, and after the classification is completed, a third blood vessel segmentation result of the vascular transition region can be obtained based on all pixels classified as blood vessels. The classification method of each pixel in the vascular transition region includes but is not limited to a statistical model classification algorithm, such as a Gaussian mixture model classification algorithm, a Gaussian-Rayleigh mixture model classification algorithm, and the like.
[0112] In this embodiment, the classification result of the vascular transition region is obtained by classifying the vascular transition region; based on the classification result that meets the preset result, the third vascular segmentation result of the vascular transition region is obtained, and the blood vessels in the vascular transition region can be finely segmented to improve the segmentation accuracy and avoid the phenomenon of insufficient segmentation of blood vessels in the vascular transition region.
[0113] In one embodiment, the step of obtaining a third blood vessel segmentation result of the blood vessel transition region according to the classification result that meets the preset result may specifically include: obtaining an initial blood vessel segmentation result of the blood vessel transition region according to the classification result that meets the preset result; and performing contour updating processing on the initial blood vessel segmentation result to obtain a third blood vessel segmentation result.
[0114] In a specific implementation, the terminal may determine all pixels classified as blood vessels as the initial blood vessel segmentation result of the blood vessel transition area, and perform contour optimization processing on the initial blood vessel segmentation result, including eliminating jagged points, noise points and other optimization processing, to obtain a third blood vessel segmentation result of the blood vessel transition area.
[0115] In this embodiment, an initial blood vessel segmentation result of the blood vessel transition area is obtained according to a classification result that meets a preset result; and a contour updating process is performed on the initial blood vessel segmentation result to obtain a third blood vessel segmentation result, thereby improving the contour of the third blood vessel segmentation result and further improving the accuracy of the third blood vessel segmentation result.
[0116] In order to facilitate those skilled in the art to have a deeper understanding of the embodiments of the present application, a specific example will be described below.
[0117] In order to solve the problem of under-segmentation at the edge of brain tissue blood vessels during cerebrovascular image segmentation, such as Figure 4 As shown, this application proposes a cerebral blood vessel segmentation method, and the specific steps include:
[0118] 1. Load cerebrovascular images, including but not limited to MRA or CTA images, and record the cerebrovascular image as Image1;
[0119] 2. Use cerebrovascular segmentation algorithms (including but not limited to deep learning algorithms and traditional segmentation algorithms) to segment the cerebrovascular image Image1 and obtain the following: Figure 5 The rough segmentation result of the cerebral vascular image is shown as Vessel1;
[0120] 3. Use the brain tissue segmentation algorithm to segment the brain tissue area of the cerebrovascular image Image1, and obtain the following Figure 6 The mask of the brain tissue area is shown as Brain-Mask;
[0121] 4. By performing AND operation on the rough segmentation result Vessel1 of the cerebral vascular image and the mask Brain-Mask, we can obtain Figure 7 The rough segmentation result of blood vessels in the brain tissue area is shown as Vessel2;
[0122] 5. Perform fine segmentation on the vascular transition region of Vessel2, the rough segmentation result of the blood vessels in the brain tissue region, to obtain Vessel3, the fine segmentation result of the blood vessels in the vascular transition region;
[0123] like Figure 8 As shown, the fine segmentation process of the vascular transition region may include:
[0124] (1) For Fig. 9 The rough segmentation result Vessel2 of the brain tissue area shown in FIG. 1 is morphologically expanded to obtain the difference between the expanded rough segmentation result Vessel2′ and Vessel2, and the mask of the difference is ANDed with the cerebral vascular image Image1 to obtain the following: Fig.10 The vascular transition zone Zone 1 of the brain tissue is shown;
[0125] Alternatively, the cerebrovascular image Image1 is ANDed with the mask Brain-Mask of the brain tissue region to obtain a vascular image Image2 containing only the brain tissue region, that is, the image outside the brain tissue is set to the background, the pixel value is 0, and the image inside the brain tissue is set to the foreground, and the pixel value is the pixel value of the cerebrovascular image Image1; the vascular rough segmentation result Vessel2 of the brain tissue region is removed from the vascular image Image2 containing only the brain tissue region to obtain a vascular image Image3 of the brain tissue region with the vascular removed. Specifically, since there is no need to segment Vessel2 again, the vascular points at the position of Vessel2 in Image2 can be set to the background; the expanded vascular rough segmentation result Vessel2' is intersected with the vascular image Image3 of the brain tissue region with the vascular removed to obtain a vascular transition region Zone1 of the brain tissue;
[0126] (2) The statistical model classification algorithm is used to classify the vascular transition region Zone1, and the initial vascular segmentation result Vessel21 of the vascular transition region Zone1 is obtained, as shown in FIG. Fig.11 As shown; wherein, the statistical model classification algorithm can adopt a Gaussian mixture model, which uses two Gaussian models for mixed modeling. When used for blood vessel classification, the blood vessel transition area can be further subdivided into background and blood vessels. This is only an example for illustration. The classification method of the blood vessel transition area is not limited to the Gaussian mixture classification algorithm, and can also be a Gaussian-Rayleigh mixture classification algorithm;
[0127] (3) Optimize the contour of the initial vessel segmentation result Vessel21 in the vessel transition region, including but not limited to removing jagged points and noise points, and obtain the following: Fig.12 The vessel segmentation result of the vascular transition area shown is Vessel3;
[0128] 6. If Fig.13 As shown in the figure, the rough segmentation result Vessel1 of the cerebrovascular image and the fine segmentation result Vessel3 of the vascular transition area are combined to finally obtain the following Fig.14 The fine brain blood vessel Vessel4 is shown.
[0129] The above-mentioned cerebral blood vessel segmentation method can automatically realize the extraction of blood vessels inside and on the surface of brain tissue in MRA or CTA images. This method uses a deep learning method to roughly segment the blood vessels inside the brain tissue, and then finely segment the surrounding areas of the roughly segmented blood vessels, effectively avoiding the under-segmentation and over-segmentation problems of cerebral blood vessel extraction, and effectively solving the under-segmentation problem of image edges when deep learning performs blood vessel segmentation. Compared with existing methods, this method has certain adaptability, can accurately segment blood vessels inside brain tissue, is suitable for scanning images of different devices, and can ensure the integrity of brain tissue blood vessel extraction.
[0130] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.
[0131] Based on the same inventive concept, the embodiment of the present application also provides a medical image segmentation device for implementing the medical image segmentation method involved above. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above method, so the specific limitations in one or more medical image segmentation device embodiments provided below can refer to the limitations of the medical image segmentation method above, and will not be repeated here.
[0132] In one embodiment, Fig.15 As shown, a medical image segmentation device is provided, comprising: a first segmentation module 210, a second segmentation module 220, a third segmentation module 230 and a merging processing module 240, wherein:
[0133] A first segmentation module 210, used to determine a first blood vessel segmentation result of the image to be segmented;
[0134] A second segmentation module 220, configured to determine a second blood vessel segmentation result of a target part in the image to be segmented according to the first blood vessel segmentation result;
[0135] A third segmentation module 230 is used to determine a vascular transition region of the target part according to the second vascular segmentation result, and determine a third vascular segmentation result of the vascular transition region;
[0136] The merging processing module 240 is used to merge the first blood vessel segmentation result and the third blood vessel segmentation result to obtain a target blood vessel segmentation result of the image to be segmented.
[0137] In one embodiment, the third segmentation module 230 is further used to expand the second blood vessel segmentation result to obtain an expanded second blood vessel segmentation result; and determine the blood vessel transition region of the target site according to the expanded second blood vessel segmentation result.
[0138] In one embodiment, the third segmentation module 230 is further used to obtain a first mask image of the blood vessel transition region according to the difference between the expanded second blood vessel segmentation result and the second blood vessel segmentation result; and determine the blood vessel transition region of the target part from the image to be segmented according to the first mask image.
[0139] In one embodiment, the second segmentation module 220 is further used to segment the target part in the image to be segmented to obtain a second mask image of the target part; and determine the second blood vessel segmentation result of the target part from the first blood vessel segmentation result according to the second mask image.
[0140] In one embodiment, the third segmentation module 230 is further used to determine a target part image from the image to be segmented according to the second mask image; remove the second blood vessel segmentation result from the target part image to obtain the target part image with the blood vessels removed; and obtain the blood vessel transition region of the target part according to the intersection of the expanded second blood vessel segmentation result and the target part image with the blood vessels removed.
[0141] In one embodiment, the third segmentation module 230 is further used to classify the vascular transition region to obtain a classification result of the vascular transition region; and obtain the third vascular segmentation result of the vascular transition region according to the classification result that meets the preset result.
[0142] In one embodiment, the third segmentation module 230 is further used to obtain an initial blood vessel segmentation result of the blood vessel transition region according to the classification result that meets the preset result; and perform contour update processing on the initial blood vessel segmentation result to obtain the third blood vessel segmentation result.
[0143] Each module in the above-mentioned medical image segmentation device can be implemented in whole or in part by software, hardware or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.
[0144] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig.16As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a medical image segmentation method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.
[0145] Those skilled in the art will understand that Fig.16 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0146] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.
[0147] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.
[0148] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.
[0149] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.
[0150] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.
[0151] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0152] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.
Claims
1. A medical image segmentation method, characterized in that: The method comprises: Determine a first blood vessel segmentation result of the image to be segmented; Determining a second blood vessel segmentation result of a target portion in the image to be segmented according to the first blood vessel segmentation result; determining a blood vessel transition region of the target site according to the second blood vessel segmentation result, and determining a third blood vessel segmentation result of the blood vessel transition region; The first blood vessel segmentation result and the third blood vessel segmentation result are combined to obtain a target blood vessel segmentation result of the image to be segmented.
2. The method according to claim 1, characterized in that: The step of determining the blood vessel transition region of the target part according to the second blood vessel segmentation result includes: performing expansion processing on the second blood vessel segmentation result to obtain an expanded second blood vessel segmentation result; The blood vessel transition region of the target site is determined according to the dilated second blood vessel segmentation result.
3. The method according to claim 2, characterized in that The step of determining the blood vessel transition region of the target site according to the segmentation result of the expanded second blood vessel comprises: obtaining a first mask image of the blood vessel transition region according to a difference between the dilated second blood vessel segmentation result and the second blood vessel segmentation result; The blood vessel transition region of the target part is determined from the image to be segmented according to the first mask image.
4. The method according to claim 2, characterized in that: The step of determining a second blood vessel segmentation result of a target part in the image to be segmented according to the first blood vessel segmentation result includes: Performing segmentation processing on the target part in the image to be segmented to obtain a second mask image of the target part; The second blood vessel segmentation result of the target part is determined from the first blood vessel segmentation result according to the second mask image.
5. The method according to claim 4, characterized in that The step of determining the blood vessel transition region of the target site according to the segmentation result of the expanded second blood vessel comprises: Determining a target part image from the image to be segmented according to the second mask image; removing the second blood vessel segmentation result from the target part image to obtain a target part image with blood vessels removed; The blood vessel transition region of the target part is obtained according to the intersection of the expanded second blood vessel segmentation result and the blood vessel removed target part image.
6. The method according to claim 1, characterized in that The determining of the third blood vessel segmentation result of the blood vessel transition region includes: performing classification processing on the vascular transition region to obtain a classification result of the vascular transition region; The third blood vessel segmentation result of the blood vessel transition area is obtained according to the classification result that meets the preset result.
7. The method according to claim 6, characterized in that The step of obtaining the third blood vessel segmentation result of the blood vessel transition region according to the classification result that meets the preset result includes: Obtaining an initial blood vessel segmentation result of the blood vessel transition region according to the classification result that meets the preset result; The initial blood vessel segmentation result is processed by contour updating to obtain the third blood vessel segmentation result.
8. A medical image segmentation device, characterized in that: The device comprises: A first segmentation module, used to determine a first blood vessel segmentation result of the image to be segmented; A second segmentation module, used to determine a second blood vessel segmentation result of a target part in the image to be segmented according to the first blood vessel segmentation result; a third segmentation module, configured to determine a vascular transition region of the target site according to the second vascular segmentation result, and determine a third vascular segmentation result of the vascular transition region; The merging processing module is used to merge the first blood vessel segmentation result and the third blood vessel segmentation result to obtain a target blood vessel segmentation result of the image to be segmented.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.