A blood vessel contour extraction method and device, electronic equipment and storage medium

Abstract

The application provides a blood vessel contour extraction method and device, electronic equipment and storage medium, including: traversing the cross-sectional slice images of the blood vessel wall 3D image, determining the connected regions of each cross-sectional slice image; connecting the centroids of each connected region in turn according to the blood vessel direction to determine the blood vessel centerline; selecting multiple points on the blood vessel centerline as target points; for each target point, intercepting a plane including the blood vessel wall range perpendicular to the tangent direction of the blood vessel centerline where the target point is located to obtain the blood vessel wall cross-section image corresponding to each target point; for each blood vessel wall cross-section image, according to the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross-section image, performing contour point extraction to obtain the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set, and performing curve fitting processing to determine the closed blood vessel inner diameter contour line and the blood vessel outer diameter contour line. The accuracy of the blood vessel inner and outer diameter contour extraction can be effectively improved.

Description

TECHNICAL FIELD

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

[0002] Globally, cardiovascular and cerebrovascular diseases have become one of the major diseases threatening human health, so in medical imaging, accurately segmenting blood vessels such as coronary artery blood vessels and extracting blood vessel contours are of great significance for determining coronary artery stenosis, lesions, etc.

[0003] In the prior art, intravenous ultrasound (IVUS) technology is usually used for intramural and adventitial edge detection, wherein intravenous ultrasound (IVUS) refers to a medical imaging technology that combines non-invasive ultrasound technology and invasive catheter technology, using a special catheter with an ultrasonic probe at the end.

[0004] However, since the IVUS image twists with the beating of the heart, it is not conducive to the extraction of the intima contour of the lumen, resulting in slow speed and low accuracy of the extraction of the intima contour of the lumen of the blood vessel. Therefore, how to improve the accuracy of blood vessel contour extraction is a technical problem to be solved by those skilled in the art. SUMMARY

[0005] Therefore, the purpose of the present application is to provide a blood vessel contour extraction method, device, electronic equipment and storage medium, which can effectively improve the accuracy of the extraction of the intramural and adventitial contour.

[0006] The present application provides a blood vessel contour extraction method, which comprises:

[0007] Obtaining a blood vessel wall 3D image segmented based on a semantic algorithm;

[0008] Traversing the cross-sectional slice images of the blood vessel wall 3D image to determine the connected regions of each cross-sectional slice image;

[0009] Connecting the centroids of each connected region in sequence according to the blood vessel direction to determine the blood vessel centerline;

[0010] Selecting a plurality of points on the blood vessel centerline as target points;

[0011] For each target point, a plane including the blood vessel wall range perpendicular to the tangent direction of the blood vessel centerline where the target point is located is intercepted to obtain the blood vessel wall cross-section corresponding to each target point;

[0012] For each blood vessel wall cross section image, according to the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross section image, contour point extraction is performed to obtain a blood vessel inner diameter contour point set and a blood vessel outer diameter contour point set;

[0013] Curve fitting processing is respectively performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set to determine closed blood vessel inner diameter contour lines and blood vessel outer diameter contour lines.

[0014] Optionally, the blood vessel wall 3D image is determined by the following steps:

[0015] A 3D medical image to be detected including a target blood vessel is acquired;

[0016] The 3D medical image to be detected is input into a pre-trained image segmentation model for blood vessel wall semantic labeling to obtain a semantic labeling result of the blood vessel wall;

[0017] Based on the semantic labeling result of the blood vessel wall, the blood vessel wall in the 3D medical image to be detected is segmented, and the blood vessel wall 3D image is output by the image segmentation model.

[0018] Optionally, the connected regions of each cross-sectional slice image are determined by the following steps:

[0019] For each cross-sectional slice image, it is identified whether there is an overlapping connected region in the connected region of the cross-sectional slice image; wherein the overlapping connected region is a connected region in which at least two blood vessel walls are identified as one;

[0020] Based on the morphological opening operation, the identified overlapping connected region is segmented to determine the actual connected region included in each cross-sectional slice image.

[0021] Optionally, the contour point extraction for each blood vessel wall cross section image based on the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross section image includes:

[0022] For each blood vessel wall cross section image, a target point of the blood vessel wall cross section image is taken as a starting point, and 360° rays are emitted at a preset interval angle;

[0023] The coordinates at which each ray first reaches the blood vessel wall segmentation result are taken as blood vessel inner diameter contour points, and the coordinates at which each ray first leaves the blood vessel wall segmentation result are taken as blood vessel outer diameter contour points;

[0024] Based on all the determined blood vessel inner diameter contour points, a blood vessel inner diameter contour point set is determined, and based on all the determined blood vessel outer diameter contour points, a blood vessel outer diameter contour point set is determined.

[0025] Optionally, the method further comprises: for each blood vessel wall cross-section image, performing contour point extraction according to a blood vessel wall segmentation result of a ring-shaped tubular hollow structure included in the blood vessel wall cross-section image, to obtain a blood vessel inner diameter contour point set and a blood vessel outer diameter contour point set, comprising:

[0026] For the blood vessel wall segmentation result in each blood vessel wall cross-section image, the blood vessel in the middle of the blood vessel wall segmentation result is filled based on morphology, to obtain a blood vessel inner wall segmentation result and a blood vessel outer wall segmentation result;

[0027] According to an edge extraction algorithm, edge point extraction is performed on the blood vessel inner wall segmentation result of each blood vessel wall cross-section image, to obtain a blood vessel inner diameter contour point set;

[0028] According to an edge extraction algorithm, edge point extraction is performed on the blood vessel outer wall segmentation result of each blood vessel wall cross-section image, to obtain a blood vessel outer diameter contour point set.

[0029] Optionally, the curve fitting processing on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set respectively comprises:

[0030] The interpolation processing is performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set respectively, the number of coordinate points in the two contour point sets is increased, and the curve fitting processing is performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set after the interpolation processing respectively.

[0031] The application further provides a blood vessel contour extraction device, the extraction device comprising:

[0032] An acquisition module is configured to acquire a blood vessel wall 3D image segmented based on a semantic algorithm;

[0033] A traversal module is configured to traverse cross-sectional slice images of the blood vessel wall 3D image, and determine a connected region of each cross-sectional slice image;

[0034] A connection module is configured to connect, according to a blood vessel direction, a center of mass of each connected region in sequence, and determine a blood vessel center line;

[0035] A selection module is configured to select a plurality of points on the blood vessel center line as target points;

[0036] A cutting module is configured to, for each target point, cut a plane including a blood vessel wall range in a direction perpendicular to a tangent direction of a blood vessel center line where the target point is located, to obtain a blood vessel wall cross-section image corresponding to each target point respectively;

[0037] An extraction module is configured to, for each blood vessel wall cross-section image, perform contour point extraction according to a blood vessel wall segmentation result of a ring-shaped tubular hollow structure included in the blood vessel wall cross-section image, to obtain a blood vessel inner diameter contour point set and a blood vessel outer diameter contour point set;

[0038] a processing module, configured to perform curve fitting processing on the set of vessel inner diameter profile points and the set of vessel outer diameter profile points respectively, to determine closed vessel inner diameter profile lines and vessel outer diameter profile lines.

[0039] Optionally, the extraction device further comprises a determination module, configured to:

[0040] obtain a 3D medical image to be detected including a target blood vessel;

[0041] input the 3D medical image to be detected into a pre-trained image segmentation model for semantic labeling of a blood vessel wall, to obtain a semantic labeling result of the blood vessel wall;

[0042] segment the blood vessel wall in the 3D medical image to be detected based on the semantic labeling result of the blood vessel wall, and output a blood vessel wall 3D image by the image segmentation model.

[0043] Optionally, the traversal module is further configured to:

[0044] for each cross-sectional slice image, identify whether there is an overlapping connected region in the connected region in the cross-sectional slice image; wherein the overlapping connected region is a connected region in which at least two blood vessel walls are identified as one;

[0045] perform segmentation processing on the identified overlapping connected region based on a morphological opening operation, to determine the connected region actually included by each cross-sectional slice image.

[0046] Optionally, when the extraction module is configured to, for each blood vessel wall cross-sectional image, perform profile point extraction based on the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross-sectional image, to obtain a set of vessel inner diameter profile points and a set of vessel outer diameter profile points, the extraction module is configured to:

[0047] for each blood vessel wall cross-sectional image, emit 360° rays at a preset interval angle with a target point of the blood vessel wall cross-sectional image as a starting point;

[0048] take the coordinates at which each ray first reaches the blood vessel wall segmentation result as a vessel inner diameter profile point, and take the coordinates at which each ray first leaves the blood vessel wall segmentation result as a vessel outer diameter profile point;

[0049] determine a set of vessel inner diameter profile points based on all the determined vessel inner diameter profile points, and determine a set of vessel outer diameter profile points based on all the determined vessel outer diameter profile points.

[0050] Optionally, when the extraction module is configured to extract the inner-diameter profile point set and the outer-diameter profile point set from the blood vessel wall segmentation result of the annular tube-shaped hollow structure included in each blood vessel wall cross-section image, the extraction module is configured to:

[0051] fill the blood vessel in the middle of the blood vessel wall segmentation result based on morphology, to obtain the inner-wall segmentation result and the outer-wall segmentation result of the blood vessel;

[0052] extract edge points from the inner-wall segmentation result of each blood vessel wall cross-section image according to an edge extraction algorithm, to obtain the inner-diameter profile point set;

[0053] extract edge points from the outer-wall segmentation result of each blood vessel wall cross-section image according to an edge extraction algorithm, to obtain the outer-diameter profile point set.

[0054] Optionally, when the processing module is configured to perform curve fitting processing on the inner-diameter profile point set and the outer-diameter profile point set respectively, the processing module is configured to:

[0055] perform interpolation processing on the inner-diameter profile point set and the outer-diameter profile point set respectively, to increase the number of coordinate points in the two profile point sets, and perform curve fitting processing on the inner-diameter profile point set and the outer-diameter profile point set after interpolation processing respectively.

[0056] The embodiments of the present application further provide an electronic device, which comprises a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicate through the bus, and the machine readable instructions are executed by the processor to perform the steps of the extraction method as described above.

[0057] The embodiments of the present application further provide a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to perform the steps of the extraction method as described above.

[0058] The embodiment of the application provides a blood vessel contour extraction method, device, electronic equipment and storage medium, which comprises the following steps: traversing cross-section slice images of a blood vessel wall 3D image, determining a connected region of each cross-section slice image; connecting the center of mass of each connected region in sequence according to the blood vessel direction to determine a blood vessel center line; selecting a plurality of points on the blood vessel center line as target points; for each target point, intercepting a plane including the blood vessel wall range in the direction perpendicular to the tangent direction of the blood vessel center line where the target point is located to obtain a blood vessel wall cross-section image corresponding to each target point; for each blood vessel wall cross-section image, performing contour point extraction according to the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross-section image to obtain a blood vessel inner diameter contour point set and a blood vessel outer diameter contour point set, and performing curve fitting processing to determine a closed blood vessel inner diameter contour line and a blood vessel outer diameter contour line.

[0059] In this way, by the technical scheme provided by the application, the blood vessel center line is determined based on the center of mass of the connected region in the cross-section slice image, the speed and efficiency of determining the blood vessel center line are improved, the blood vessel inner wall segmentation result and the blood vessel outer wall segmentation result are obtained by adopting the morphological filling operation, the accuracy of determining the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set is improved, and the accuracy of blood vessel contour extraction is further improved.

[0060] In order to make the above-mentioned purpose, characteristics and advantages of the application more obvious and easy to understand, the following preferred embodiments are described in detail below, and the accompanying drawings are described as follows. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical scheme of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0062] Figure 1 The flow chart of the blood vessel contour extraction method provided by the embodiments of the application;

[0063] Figure 2 The structural schematic diagram of the blood vessel wall 3D image provided by the application;

[0064] Figure 3 The process schematic diagram of the post-processing of the blood vessel wall 3D image provided by the application;

[0065] Figure 4 The structural schematic diagram of the blood vessel center line provided by the application;

[0066] Figure 5 The structural schematic diagram of the segmentation result provided by the application;

[0067] Figure 6 A structural schematic diagram of a blood vessel inner diameter profile provided by the present application;

[0068] Figure 7 A structural schematic diagram of a blood vessel profile extraction device provided by an embodiment of the present application;

[0069] Figure 8 A structural schematic diagram of a blood vessel profile extraction device provided by an embodiment of the present application;

[0070] Figure 9 A structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0071] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. The components of the embodiments of the present application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, every other embodiment obtained by a person skilled in the art without creative work falls within the scope of the present application.

[0072] Globally, cardiovascular and cerebrovascular diseases have become one of the major diseases threatening human health. Therefore, in medical imaging, it is of great significance to accurately segment blood vessels such as coronary artery blood vessels and extract blood vessel profiles for determining the stenosis, lesions, etc. of the coronary artery.

[0073] In the prior art, intravenous ultrasound (IVUS) technology is usually used for intramural and adventitial edge detection. Intravenous ultrasound (IVUS) refers to a medical imaging technology that combines non-invasive ultrasound technology and invasive catheter technology, using a special catheter with an ultrasonic probe at the end.

[0074] However, since the IVUS image twists with the beating of the heart, it is not conducive to the extraction of the intimal profile of the lumen, resulting in a slow speed and low accuracy of the extraction of the intimal profile of the lumen. Therefore, how to improve the accuracy of the extraction of the blood vessel profile is a technical problem to be solved by those skilled in the art.

[0075] Based on this, the embodiment of the present application provides a blood vessel contour extraction method, device, electronic equipment and storage medium, which can effectively improve the accuracy of blood vessel contour extraction.

[0076] Please refer to Figure 1 , Figure 1 The flowchart of the blood vessel contour extraction method provided by the embodiment of the present application is shown in FIG. 1. Figure 1 As shown in FIG. 1, the extraction method provided by the embodiment of the present application comprises the following steps.

[0077] S101, obtaining a blood vessel wall 3D image segmented based on a semantic algorithm.

[0078] Here, the blood vessel wall 3D image is a 3D image in which interfering tissues such as muscles and bones are removed and only the blood vessel wall structure is retained.

[0079] In an embodiment, the blood vessel wall 3D image is determined by the following steps.

[0080] S1011, obtaining a 3D medical image to be detected including a target blood vessel.

[0081] Here, the 3D medical image to be detected can be a CT, MRI or the like. When the target blood vessel is a coronary blood vessel, the 3D medical image to be detected can be a heart 3D medical image.

[0082] S1012, inputting the 3D medical image to be detected into a pre-trained image segmentation model to perform blood vessel wall semantic labeling, to obtain a semantic labeling result of the blood vessel wall.

[0083] Here, the blood vessel wall semantic labeling of the 3D medical image to be detected is to label the voxels corresponding to the blood vessel wall.

[0084] The image segmentation model can be constructed by the following steps: obtaining a plurality of original medical image data, the original medical image data being data added with blood vessel wall labeling; inputting the original medical image data into a deep neural network to obtain a prediction result of the blood vessel wall; comparing the prediction result with the real blood vessel wall labeling to determine a loss function value, and feeding back the determined loss function value to the deep neural network; updating the deep neural network based on the feedback loss function value to evolve it in the direction of reducing the prediction error, and finally obtaining the image segmentation model.

[0085] The deep neural network can be a classical medical semantic segmentation network such as U-Net, V-Net, etc.; and the loss function can be BCE, cross-entropy, etc.

[0086] S1013, based on the semantic annotation result of the blood vessel wall, segmenting the blood vessel wall in the to-be-detected 3D medical image, and outputting a blood vessel wall 3D image by the image segmentation model.

[0087] For example, refer to Figure 2 , Figure 2 A structural schematic diagram of a blood vessel wall 3D image provided by the present application.

[0088] In addition, before step S102 is performed, the blood vessel wall 3D image can also be post-processed to eliminate the influence of some unnecessary blood vessels on the accuracy of subsequent extraction of the blood vessel contour. For example, refer to Figure 3 , Figure 3 A process schematic diagram of post-processing of a blood vessel wall 3D image provided by the present application. As Figure 3 shown, here the post-processing process is based on the maximum connected domain, first obtaining the binary image of the preliminary predicted blood vessel wall segmentation result (Pf) shown in the left figure (element 1 is the peripheral blood vessel and element 0 is the background); then based on Figure 3 the left figure, performing maximum connected region processing to filter out non-maximum connected regions to obtain the maximum connected region, and finally using Figure 3 the left figure to multiply the maximum connected region to obtain the right figure in Figure 3 , completing the post-processing. Figure 3

[0089] S102, traversing the cross-sectional slice images of the blood vessel wall 3D image to determine the connected region of each cross-sectional slice image.

[0090] The cross section can also be referred to as a horizontal plane, and the cross section is usually a plane determined by cutting the 3D image from top to bottom.

[0091] Here, the blood vessel wall 3D image includes a plurality of cross-sectional slice images, and the number of cross-sectional slice images can be accurately determined by the device used to obtain the to-be-detected 3D medical image, or the number of cross-sectional slice images that can be segmented from the blood vessel wall 3D image can be artificially set.

[0092] Among them, there can be at least one connected region in the cross-sectional slice image, or there can be no connected region.

[0093] In an embodiment provided by the present application, the connected region of each cross-sectional slice image is determined by the following steps: for each cross-sectional slice image, identifying whether the connected region in the cross-sectional slice image has an overlapping connected region; based on the morphological opening operation, performing segmentation processing on the identified overlapping connected region to determine the actual connected region included by each cross-sectional slice image.

[0094] ​Here, the overlapping connected region is a connected region in which connected regions of at least two vessel walls are identified as one.

[0095] It should be noted that, for the cross-sectional slice image of the vessel wall at the bifurcation, when the connected region identification is performed, the connected regions of adjacent vessel walls are usually identified as one connected region, thereby affecting the accuracy of subsequent determination of the vessel centerline, and therefore the connected region overlapping identification needs to be performed, so as to perform subsequent segmentation processing.

[0096] For example, whether it is an overlapping connected region can be determined by identifying whether the connected region has a concave hull, wherein the connected region having the concave hull is determined as the overlapping connected region, and the segmentation processing is performed.

[0097] The segmentation processing of the identified overlapping connected region based on the morphological opening operation includes: performing an opening operation processing on the cross-sectional slice image corresponding to the overlapping connected region based on the morphological operation; and performing a connected region identification processing on the cross-sectional slice image after the opening operation processing, to determine the connected region included in the cross-sectional slice image. The opening operation processing is a corrosion processing followed by an inflation processing on the cross-sectional slice image.

[0098] S103, sequentially connecting the center of mass of each connected region according to the vessel direction to determine the vessel centerline.

[0099] Here, the center of mass of the connected region is also the center of mass of the vessel wall, which can be used as the center point of the vessel wall. The center of mass of each connected region is determined, that is, the center point set of the vessel wall is determined.

[0100] It should be noted that the vessel wall surrounds the blood vessel, and the center of the blood vessel is basically coincident with the center of the vessel wall, and therefore the vessel centerline can be determined according to the center point set of the vessel wall.

[0101] For example, please refer to Figure 4 , Figure 4 A structural schematic diagram of a vessel centerline provided in the present application.

[0102] S104, selecting a plurality of points on the vessel centerline as target points.

[0103] Here, the determined center of mass can be directly used as the selected plurality of target points; or a plurality of points can be extracted from the vessel centerline at equal intervals as the target points; or the extraction rule of the target points can be specified based on requirements, so as to obtain a plurality of target points.

[0104] S105, for each target point, intercepting a plane including the range of the vessel wall perpendicular to the tangent direction of the vessel centerline where the target point is located, to obtain a vessel wall cross-sectional image corresponding to each target point.

[0105] Here, the cross-sectional view of the blood vessel wall is a tangent plane of the blood vessel at the target point location. This tangent plane is perpendicular to the tangent direction of the center line at this location, and this tangent plane is image data of a certain size (including the range of the blood vessel wall).

[0106] Each cross-sectional image of the vessel wall includes a segmentation result of the vessel wall. For an example, please refer to [link to example]. Figure 5 , Figure 5 This is a schematic diagram of a segmentation result provided in this application. For example... Figure 5 As shown, Figure 5 (a) in the figure represents the segmentation result of the blood vessel wall, where the gray part represents the structure of the blood vessel wall.

[0107] S106. For each blood vessel wall cross-sectional image, based on the blood vessel wall segmentation results of the annular tubular hollow structure included in the blood vessel wall cross-sectional image, extract the contour points to obtain the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set.

[0108] Here, each cross-sectional view of the blood vessel wall can yield a set of inner diameter contour points and a set of outer diameter contour points. Based on the set of inner diameter contour points and the set of outer diameter contour points determined by each cross-sectional view of the blood vessel, the inner diameter contour point set and the outer diameter contour point set of the blood vessel are obtained.

[0109] In one embodiment, step S106, for each blood vessel wall cross-sectional image, involves extracting contour points based on the segmentation results of the blood vessel wall including the annular tubular hollow structure in the cross-sectional image, to obtain a set of contour points for the inner diameter of the blood vessel and a set of contour points for the outer diameter of the blood vessel, including:

[0110] S10611. For each cross-sectional view of the blood vessel wall, starting from the target point of the cross-sectional view of the blood vessel wall, emit rays at preset intervals of 360°.

[0111] S10612. Take the coordinates of each ray when it first reaches the segmentation result of the blood vessel wall as the inner diameter contour point of the blood vessel, and take the coordinates of each ray when it first leaves the segmentation result of the blood vessel wall as the outer diameter contour point of the blood vessel.

[0112] S10613. Determine the set of inner diameter contour points of blood vessels based on all determined inner diameter contour points, and determine the set of outer diameter contour points of blood vessels based on all determined outer diameter contour points.

[0113] Here, steps S10611 to S10613 describe the implementation process of determining the inner diameter contour point set and outer diameter contour point set of blood vessels based on the ray method.

[0114] In another embodiment, in step S106, for each blood vessel wall cross-section image, according to the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross-section image, contour point extraction is performed to obtain a blood vessel inner diameter contour point set and a blood vessel outer diameter contour point set, including:

[0115] S10621, for the blood vessel wall segmentation result in each blood vessel wall cross-section image, the blood vessel in the middle of the blood vessel wall is filled based on morphology to obtain a blood vessel inner wall segmentation result and a blood vessel outer wall segmentation result;

[0116] S10622, according to an edge extraction algorithm, edge point extraction is performed on the blood vessel inner wall segmentation result of each blood vessel wall cross-section image to obtain a blood vessel inner diameter contour point set;

[0117] S10623, according to an edge extraction algorithm, edge point extraction is performed on the blood vessel outer wall segmentation result of each blood vessel wall cross-section image to obtain a blood vessel outer diameter contour point set.

[0118] Here, steps S10621 to S10623 describe the implementation process of determining the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set based on morphological filling operation and edge extraction method.

[0119] Please continue to refer to Figure 5 , Figure 5 (b) is the blood vessel inner wall segmentation result determined based on the morphological filling operation, Figure 5 (c) is the blood vessel outer wall segmentation result determined based on the morphological filling operation.

[0120] It should be noted that the reason why the filling and segmentation are needed to determine the blood vessel inner wall segmentation result and the blood vessel outer wall segmentation result is that if the edge extraction method is directly used for contour point extraction based on the blood vessel wall segmentation result, when the inner and outer diameter contours have overlapping points, the outer diameter contour points will be extracted when the inner diameter contour points are extracted, or the inner diameter contour points will be extracted when the outer diameter contour points are extracted, resulting in contour point extraction error.

[0121] S107, curve fitting processing is respectively performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set to determine closed blood vessel inner diameter contour lines and blood vessel outer diameter contour lines.

[0122] In one embodiment, the curve fitting processing on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set respectively includes: interpolation processing is respectively performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set to increase the number of coordinate points in the two contour point sets, and curve fitting processing is respectively performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set after interpolation processing.

[0123] Here, the contour point set is interpolated to be closer to the real shape of the contour and to increase the contour points to enrich the contour details.

[0124] For example, refer to Fig. 6, Figure 6 A structure diagram of the intravascular diameter contour line is provided in the present application. For better observation, only the intravascular diameter contour line is displayed here, and the display of the extravascular diameter contour line is similar to that of the intravascular diameter contour line.

[0125] In addition, after the intravascular diameter contour line and the extravascular diameter contour line are determined, the intravascular contour surface and the extravascular contour surface can also be determined through curved surface smoothing processing.

[0126] The method for extracting the blood vessel contour provided in the embodiments of the present application comprises the following steps: traversing the cross-sectional slice images of the blood vessel wall 3D image to determine the connected regions of each cross-sectional slice image; connecting the centroids of each connected region in turn according to the blood vessel direction to determine the blood vessel centerline; selecting a plurality of points on the blood vessel centerline as target points; for each target point, intercepting a plane including the blood vessel wall range in the direction perpendicular to the tangent direction of the blood vessel centerline where the target point is located to obtain the blood vessel wall cross-sectional image corresponding to each target point; for each blood vessel wall cross-sectional image, performing contour point extraction according to the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross-sectional image to obtain the intravascular diameter contour point set and the extravascular diameter contour point set, and performing curve fitting processing to determine the closed intravascular diameter contour line and the extravascular diameter contour line.

[0127] In this way, by using the technical solution provided in the present application, the blood vessel centerline is determined based on the centroid of the connected region in the cross-sectional slice image, the speed and efficiency of determining the blood vessel centerline are improved, the blood vessel inner wall segmentation result and the blood vessel outer wall segmentation result are obtained by using the morphological filling operation, the accuracy of determining the intravascular diameter contour point set and the extravascular diameter contour point set is improved, and the accuracy of extracting the intravascular and extravascular diameter contour is further improved.

[0128] For example, refer to Fig. 6, Figure 7 , Figure 8 , Figure 7 Fig. 1 is a structure diagram of a blood vessel contour extraction device provided in the embodiments of the present application. Figure 8 Fig. 2 is another structure diagram of the blood vessel contour extraction device provided in the embodiments of the present application. As shown in Fig. 2, Figure 7 The extraction device 700 comprises:

[0129] The acquisition module 710 is configured to acquire the blood vessel wall 3D image segmented based on a semantic algorithm.

[0130] The traversal module 720 is configured to traverse the cross-sectional slice images of the blood vessel wall 3D image to determine the connected regions of each cross-sectional slice image.

[0131] The connection module 730 is used to connect the centroids of each connected region sequentially according to the direction of the blood vessel to determine the centerline of the blood vessel.

[0132] The selection module 740 is used to select multiple points on the center line of the blood vessel as target points;

[0133] The interception module 750 is used to intercept a plane including the vessel wall area that is perpendicular to the tangent direction of the vessel centerline where the target point is located for each target point, so as to obtain the vessel wall cross-section diagram corresponding to each target point.

[0134] The extraction module 760 is used to extract contour points for each blood vessel wall cross-section image based on the blood vessel wall segmentation results of the annular tubular hollow structure included in the blood vessel wall cross-section image, so as to obtain the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set.

[0135] The processing module 770 is used to perform curve fitting processing on the set of inner diameter contour points of the blood vessel and the set of outer diameter contour points of the blood vessel respectively, and to determine the closed inner diameter contour line and outer diameter contour line of the blood vessel.

[0136] Optional, such as Figure 8 As shown, the extraction device 700 further includes a determining module 780, which is used for:

[0137] Acquire 3D medical images of the target blood vessels;

[0138] The 3D medical image to be detected is input into a pre-trained image segmentation model for semantic annotation of the blood vessel wall, and the semantic annotation result of the blood vessel wall is obtained.

[0139] Based on the semantic annotation results of the blood vessel wall, the blood vessel wall in the 3D medical image to be detected is segmented, and the image segmentation model outputs a 3D image of the blood vessel wall.

[0140] Optionally, the traversal module 720 is further configured to:

[0141] For each cross-sectional slice image, identify whether there are overlapping connected regions in the connected regions of the cross-sectional slice image; wherein, the overlapping connected region is a connected region that identifies at least two connected regions of the blood vessel wall as one;

[0142] Based on morphological opening operations, the identified overlapping connected regions are segmented to determine the actual connected regions included in each cross-sectional slice image.

[0143] Optionally, when the extraction module 760 is used to, for each blood vessel wall cross section image, perform contour point extraction according to the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross section image, to obtain the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set, the extraction module 760 is configured to:

[0144] For each blood vessel wall cross section image, a 360° ray is emitted at a preset interval angle, starting from the target point of the blood vessel wall cross section image.

[0145] The coordinate at which each ray first reaches the blood vessel wall segmentation result is taken as a blood vessel inner diameter contour point, and the coordinate at which each ray first leaves the blood vessel wall segmentation result is taken as a blood vessel outer diameter contour point.

[0146] The blood vessel inner diameter contour point set is determined based on all the determined blood vessel inner diameter contour points, and the blood vessel outer diameter contour point set is determined based on all the determined blood vessel outer diameter contour points.

[0147] Optionally, when the extraction module 760 is used to, for each blood vessel wall cross section image, perform contour point extraction according to the blood vessel wall segmentation result of the annular tubular hollow structure included in the blood vessel wall cross section image, to obtain the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set, the extraction module 760 is configured to:

[0148] For the blood vessel wall segmentation result in each blood vessel wall cross section image, the blood vessel in the middle of the blood vessel wall in the blood vessel wall segmentation result is filled based on morphology, to obtain the blood vessel inner wall segmentation result and the blood vessel outer wall segmentation result.

[0149] According to an edge extraction algorithm, edge point extraction is performed on the blood vessel inner wall segmentation result of each blood vessel wall cross section image, to obtain the blood vessel inner diameter contour point set.

[0150] According to an edge extraction algorithm, edge point extraction is performed on the blood vessel outer wall segmentation result of each blood vessel wall cross section image, to obtain the blood vessel outer diameter contour point set.

[0151] Optionally, when the processing module 770 is used to perform curve fitting processing on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set respectively, the processing module 770 is configured to:

[0152] Interpolation processing is performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set respectively, to increase the number of coordinate points in the two contour point sets, and curve fitting processing is performed on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set after interpolation processing respectively.

[0153] Please refer to Figure 9 , Figure 9 Figure 1 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. As shown in Figure 1, the electronic device comprises a processor 710, a memory 720, an extraction module 760 and a processing module 770. Figure 9As shown in FIG. 9, the electronic device 900 includes a processor 910, a memory 920 and a bus 930.

[0154] The memory 920 stores machine readable instructions executable by the processor 910, and when the electronic device 900 is running, the processor 910 communicates with the memory 920 through the bus 930. When the machine readable instructions are executed by the processor 910, the steps in the method embodiments shown above can be performed. Figures 1 to 6 The specific implementation can refer to the method embodiments, and will not be repeated here.

[0155] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program. When the computer program is run by a processor, the steps in the method embodiments shown above can be performed. The specific implementation can refer to the method embodiments, and will not be repeated here. Figures 1 to 6 The specific implementation can refer to the method embodiments, and will not be repeated here.

[0156] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0157] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented by other ways. The device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, and for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some communication interfaces, devices or units, and can be electrical, mechanical or other forms.

[0158] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the embodiment scheme.

[0159] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit.

[0160] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a nonvolatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the present application essentially or the parts of the prior art that make contributions or parts of the technical solutions can be embodied in the form of software products. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0161] Finally, it should be noted that: the above-described embodiments are only specific embodiments of the present application, which are used to illustrate the technical solutions of the present application, but not to limit them. The protection scope of the present application is not limited thereto. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can make modifications or easily think of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed by the present application, or make equivalent replacements to some of the technical features. The modifications, changes or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method of extraction of a blood vessel profile, characterized by, The extraction method comprises: obtaining a blood vessel wall 3D image segmented based on a semantic algorithm; traversing cross-sectional slice images of the blood vessel wall 3D image to determine a connected region of each cross-sectional slice image; connecting the centroids of each connected region in sequence according to the blood vessel direction to determine a blood vessel centerline; selecting a plurality of points on the blood vessel centerline as target points; for each target point, intercepting a plane including the blood vessel wall range in a direction perpendicular to the tangent of the blood vessel centerline where the target point is located to obtain a blood vessel wall cross-section corresponding to each target point; for each blood vessel wall cross-section, performing contour point extraction according to a blood vessel wall segmentation result of a ring-shaped tubular hollow structure included in the blood vessel wall cross-section to obtain a blood vessel inner diameter contour point set and a blood vessel outer diameter contour point set; the contour point extraction according to the blood vessel wall segmentation result of the ring-shaped tubular hollow structure included in each blood vessel wall cross-section to obtain the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set comprises: for the blood vessel wall segmentation result in each blood vessel wall cross-section, filling the blood vessel in the middle of the blood vessel wall segmentation result based on morphology to obtain a blood vessel inner wall segmentation result and a blood vessel outer wall segmentation result; performing edge point extraction on the blood vessel inner wall segmentation result of each blood vessel wall cross-section according to an edge extraction algorithm to obtain the blood vessel inner diameter contour point set; performing edge point extraction on the blood vessel outer wall segmentation result of each blood vessel wall cross-section according to the edge extraction algorithm to obtain the blood vessel outer diameter contour point set; respectively performing curve fitting processing on the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set to determine closed blood vessel inner diameter contour lines and blood vessel outer diameter contour lines.

2. The extraction method according to claim 1, characterized in that, The blood vessel wall 3D image is determined through the following steps: obtaining a 3D medical image to be detected including a target blood vessel; inputting the 3D medical image to be detected into a pre-trained image segmentation model for blood vessel wall semantic labeling to obtain a semantic labeling result of the blood vessel wall; segmenting the blood vessel wall in the 3D medical image to be detected based on the semantic labeling result of the blood vessel wall, and outputting the blood vessel wall 3D image from the image segmentation model.

3. The extraction method of claim 1, wherein, The connected region of each cross-sectional slice image is determined through the following steps: for each cross-sectional slice image, identifying whether there is an overlapping connected region in the connected region of the cross-sectional slice image; wherein the overlapping connected region is a connected region in which at least two blood vessel wall connected regions are identified as one; performing segmentation processing on the identified overlapping connected region based on a morphological opening operation to determine the actual connected region included in each cross-sectional slice image.

4. The extraction method of claim 1, wherein, The contour point extraction according to the blood vessel wall segmentation result of the ring-shaped tubular hollow structure included in each blood vessel wall cross-section to obtain the blood vessel inner diameter contour point set and the blood vessel outer diameter contour point set comprises: for each blood vessel wall cross-section, emitting 360° rays at a preset interval angle with the target point of the blood vessel wall cross-section as the starting point; taking the coordinates at which each ray first reaches the blood vessel wall segmentation result as the blood vessel inner diameter contour points, and taking the coordinates at which each ray first leaves the blood vessel wall segmentation result as the blood vessel outer diameter contour points; Determine a set of inner vessel diameter contour points based on the determined all inner vessel diameter contour points, and determine a set of outer vessel diameter contour points based on the determined all outer vessel diameter contour points.

5. The extraction method of claim 1, wherein, The curve fitting processing on the set of inner vessel diameter contour points and the set of outer vessel diameter contour points respectively comprises: The curve fitting processing on the set of inner vessel diameter contour points and the set of outer vessel diameter contour points respectively comprises:

6. An apparatus for extracting a blood vessel profile, characterized by comprising: The extraction device comprises: An acquisition module configured to acquire a 3D image of a vessel wall segmented based on a semantic algorithm; An iteration module configured to iterate through cross-sectional slice images of the 3D image of the vessel wall to determine a connected region of each cross-sectional slice image; A connection module configured to sequentially connect a centroid of each connected region according to a vessel direction to determine a vessel centerline; A selection module configured to select a plurality of points on the vessel centerline as target points; A cutting module configured to, for each target point, cut a plane including a range of the vessel wall in a direction perpendicular to a tangent of the vessel centerline at which the target point is located to obtain a vessel wall cross-section corresponding to each target point; An extraction module configured to, for each vessel wall cross-section, perform contour point extraction according to a vessel wall segmentation result of a ring-shaped tubular hollow structure included in the vessel wall cross-section to obtain a set of inner vessel diameter contour points and a set of outer vessel diameter contour points; when the extraction module is used to, for each vessel wall cross-section, perform contour point extraction according to a vessel wall segmentation result of a ring-shaped tubular hollow structure included in the vessel wall cross-section to obtain a set of inner vessel diameter contour points and a set of outer vessel diameter contour points, the extraction module is configured to: based on a morphological filling of a vessel in the middle of the vessel wall in the vessel wall segmentation result, obtain an inner vessel wall segmentation result and an outer vessel wall segmentation result for the vessel wall segmentation result in each vessel wall cross-section; perform edge point extraction on the inner vessel wall segmentation result of each vessel wall cross-section according to an edge extraction algorithm to obtain the set of inner vessel diameter contour points; and perform edge point extraction on the outer vessel wall segmentation result of each vessel wall cross-section according to the edge extraction algorithm to obtain the set of outer vessel diameter contour points; A processing module configured to perform curve fitting processing on the set of inner vessel diameter contour points and the set of outer vessel diameter contour points respectively to determine closed inner vessel diameter contour lines and outer vessel diameter contour lines.

7. The extraction device of claim 6, wherein The extraction device further comprises a determination module configured to: acquire a 3D medical image to be detected including a target vessel; input the 3D medical image to be detected into a pre-trained image segmentation model to perform semantic labeling on a vessel wall to obtain a semantic labeling result of the vessel wall; segment the vessel wall in the 3D medical image to be detected based on the semantic labeling result of the vessel wall, and output a 3D image of the vessel wall by the image segmentation model.

8. An electronic device, comprising: comprises: A processor, a memory, and a bus, the memory storing machine readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicating through the bus, the machine readable instructions being executed by the processor to perform the steps of the extraction method of any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that, A computer readable storage medium storing a computer program, the computer program being executed by a processor to perform the steps of the extraction method of any one of claims 1 to 5.

Images