A semi-automatic measurement method for morphological parameters of intracranial aneurysms

Abstract

The present application relates to a semi-automatic measurement method for morphological parameters of intracranial aneurysms, including: A1, based on a first mask combination of a single aneurysm, obtaining a first aneurysm diameter plane and a first blood flow direction vector; A2, based on the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, according to the calculation rules of each parameter value in the parameter information, obtaining the parameter information of the aneurysm; A3, based on the predefined aneurysm circumscribed rectangular frame condition, according to the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, obtaining the aneurysm circumscribed rectangular frame; A4, based on the first mask combination, the aneurysm circumscribed rectangular frame and the aneurysm parameter information, outputting a visualization result for doctors to view. The present method assists the segmentation of aneurysms and tumor-bearing vessels by manually adjusting tools, thereby ensuring the accuracy of parameter calculation.

Description

Technical Field

[0001] The present application belongs to the field of image processing technology, and specifically relates to a semi-automatic measurement method for morphological parameters of intracranial aneurysms. Background Art

[0002] The morphological changes of intracranial aneurysms can reflect the process of their occurrence, growth, and rupture. Aneurysm parameters are closely related to the formulation of treatment plans and the prediction of rupture risks. Therefore, the morphological measurement of intracranial aneurysms is the basis of clinical diagnosis and treatment. Accurate measurement of aneurysm morphological parameters is conducive to assessing the risk of intracranial aneurysm rupture and can assist clinicians in adjusting further treatment strategies. It has great clinical significance and social benefits in reducing disability caused by aneurysm rupture and bleeding and reducing unnecessary treatment.

[0003] At present, the measurement of intracranial aneurysms is mainly divided into traditional manual measurement and computer automatic measurement. In clinical work, manual measurement is still the most commonly used method for aneurysm morphology measurement. With the development of intelligent technology, the status of automatic measurement is gradually improving, but in clinical applications, automatic measurement has certain limitations. For example, it is affected by the complex morphology of intracranial blood vessels, and its accuracy depends on whether the aneurysm and the tumor-bearing vessel can be accurately segmented, which in turn affects the accuracy of parameter measurement.

[0004] At present, most doctors use automatic measurement methods, which require manual measurement with the help of software, which consumes a lot of doctors' time. In addition, there are some automatic measurement schemes that automatically extract and calculate parameters based on the results of the upstream aneurysm segmentation model; when the results of the upstream aneurysm segmentation model are wrong, the calculated parameters will also be wrong, and doctors cannot correct this error by manual adjustment.

[0005] In view of this, providing a more accurate method for segmenting aneurysms and parent vessels and realizing semi-automatic measurement of intracranial aneurysm morphology has become an urgent problem to be solved. Summary of the invention

[0006] 1. Technical issues to be resolved

[0007] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present application provides a semi-automatic measurement method for morphological parameters of intracranial aneurysms.

[0008] (II) Technical solution

[0009] In order to achieve the above objectives, this application adopts the following technical solutions:

[0010] In a first aspect, the present application provides a method for semi-automatically measuring morphological parameters of intracranial aneurysms, comprising:

[0011] A1. Based on the first mask combination of a single aneurysm, a first aneurysm diameter plane and a first blood flow direction vector are obtained;

[0012] The first mask combination is information of a connected domain of a single aneurysm mask and information of a connected domain of a blood vessel mask of a blood vessel to which the single aneurysm mask belongs, which are pre-extracted from the three-dimensional medical image to be analyzed;

[0013] A2. based on the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, and according to the calculation rule of each parameter value in the parameter information, obtaining parameter information of the aneurysm;

[0014] The parameter information of the aneurysm includes one or more of the following: aneurysm diameter length, maximum diameter, aneurysm height and aneurysm width;

[0015] A3, based on the predefined condition of the aneurysm circumscribed rectangular frame, according to the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, obtaining the aneurysm circumscribed rectangular frame;

[0016] A4. Based on the first mask combination, the aneurysm circumscribed rectangular frame and the parameter information of the aneurysm, a visualization result is output for the doctor to view.

[0017] Optionally, the method further includes:

[0018] A5. If the display interface of the visualization result receives an adjustment instruction triggered by the doctor;

[0019] Then, according to the adjustment parameters in the adjustment instruction, based on re-segmenting the three-dimensional medical image, a new aneurysm diameter plane, a new blood flow direction vector and an aneurysm mask in a new mask combination are obtained; the new aneurysm diameter plane is used as a second aneurysm diameter plane; the new blood flow direction vector is used as a second blood flow direction vector; and the aneurysm mask in the new mask combination is used as an aneurysm mask in the second mask combination;

[0020] For the second aneurysm diameter plane, the second blood flow direction vector and the aneurysm mask in the second mask combination corresponding to a single aneurysm, the second aneurysm diameter plane is used as the new first aneurysm diameter plane, the second blood flow direction vector is used as the new first blood flow direction vector, and the aneurysm mask in the second mask combination is used as the aneurysm mask in the new first mask combination; repeat the above processes A2 to A5.

[0021] Optionally, A1 includes:

[0022] A11: performing a dilation operation on the aneurysm mask in the first mask combination by using a cubic convolution kernel with a side length of 3 voxels, taking voxels in the dilation operation result that overlap with the blood vessel mask in the first mask combination as intersection points, and obtaining all intersection points;

[0023] A12: Calculate the geometric center of the intersection point, and use the geometric center as the estimated center of the tumor body; and select a region with the same size as the circumscribed cube of the aneurysm mask in the first mask combination, with the estimated center of the tumor body as the center;

[0024] Taking the coordinate set of all voxels on the aneurysm mask in the first mask combination in the selected area as the aneurysm point set, and taking the coordinate set of all voxels on the blood vessel mask in the first mask combination in the selected area as the blood vessel point set; acquiring the aneurysm point set and the blood vessel point set;

[0025] A13: establishing a support vector machine classification model based on the aneurysm point set and the blood vessel point set;

[0026] Fitting the support vector machine classification model to obtain a classification plane; using the obtained classification plane as a first tumor diameter plane, and using the geometric parameters of the first tumor diameter plane as first plane geometric parameters; obtaining the first tumor diameter plane and the first plane geometric parameters;

[0027] A14: taking the estimated center of the tumor as the center, select a cubic region with a side length of 15 voxels; perform three-dimensional skeletonization on the blood vessel mask in the first mask combination in the selected cubic region to obtain a skeleton point set; if a point in the skeleton point set satisfies the condition that there is only one point belonging to the skeleton point set in the 26 neighborhoods of the point, then take the point as the starting point; obtain all starting points;

[0028] A point is randomly selected from the acquired starting points as a first starting point, the first starting point is set as a current point, and a point which is closest to the current point and has not been the current point is set as a new current point; the current point is searched in the skeleton point set until the current point is another starting point;

[0029] The other starting point is used as the second starting point; the projection of the vector formed by the first starting point and the second starting point on the first tumor diameter plane is used as the first blood flow direction vector to obtain the first blood flow direction vector.

[0030] Optionally, A2 includes:

[0031] A21: based on the first plane geometric parameters, calculate the distance between each voxel on the aneurysm mask in the first mask combination and the first aneurysm diameter plane, and take the set of points whose distance is less than one voxel unit as the first aneurysm diameter point set; obtain the first aneurysm diameter point set;

[0032] A22: traverse all points on the aneurysm mask in the first mask combination, and if there is a point on the aneurysm mask that does not belong to the first mask combination within the 18 neighborhood of the current traversal point, take the current traversal point as an aneurysm surface point, and take the set of all aneurysm surface points as an aneurysm surface point set; obtain the aneurysm surface point set;

[0033] A23: traversing any two point combinations in the first tumor diameter point set; calculating the Euclidean distance between the two points in all point set combinations, and calculating the absolute value of the cosine of the angle between the vector formed by the two points in all point set combinations and the first blood flow direction vector; removing the point set combinations whose absolute value of the cosine of the angle is less than 0.9, to obtain the second tumor diameter point set;

[0034] Based on the second tumor path point set, the maximum Euclidean distance between two points in the point set combination is screened out, and the maximum Euclidean distance is used as the tumor path length; the two points corresponding to the tumor path length are used as the tumor path endpoints, and the midpoint of the tumor path endpoints is used as the tumor path midpoint; and the tumor path length, the tumor path endpoints and the tumor path midpoint are obtained;

[0035] A24: Calculate the Euclidean distances between all points in the aneurysm surface point set and the midpoint of the aneurysm diameter, take the point farthest from the midpoint of the aneurysm diameter as the maximum diameter point, take the Euclidean distance between the maximum diameter point and the midpoint of the aneurysm diameter as the maximum diameter; take the vector formed by the maximum diameter point and the midpoint of the aneurysm diameter as the maximum diameter direction vector; obtain the maximum diameter point, maximum diameter and maximum diameter direction vector;

[0036] A25: for the points in the aneurysm surface point set, delete the points that are not on the same side of the aneurysm diameter plane as the maximum diameter point;

[0037] A26: Calculate the Euclidean distance between all points in the aneurysm surface point set and the first aneurysm diameter plane, take the point farthest from the first aneurysm diameter plane as the aneurysm height point, take the Euclidean distance between the aneurysm height point and the first aneurysm diameter plane as the aneurysm body height; take the projection point of the aneurysm height point on the first aneurysm diameter plane as the aneurysm height projection point, and take the projection point as the aneurysm height projection point; obtain the aneurysm height point, aneurysm body height and aneurysm height projection point;

[0038] A27: traverse the point set combination of any two points in the aneurysm surface point set; calculate the Euclidean distance between the two points in all point set combinations, calculate the absolute value of the cosine of the angle between the vector formed by the two points in all point set combinations and the first blood flow direction vector; remove the point set combination whose absolute value of the cosine of the angle is less than 0.3, and obtain the third aneurysm diameter point set;

[0039] For the third tumor diameter point set, select the maximum Euclidean distance between two points in the point set combination, use the maximum Euclidean distance as the tumor diameter width, and use the two points corresponding to the tumor diameter width as the tumor width endpoints; obtain the tumor diameter width and the tumor width endpoints;

[0040] A28: taking the aneurysm diameter length, maximum diameter, aneurysm height and aneurysm width as parameter information of a single aneurysm, and merging the parameter information of the single aneurysm into a parameter information list of the single aneurysm; and obtaining the parameter information list of the single aneurysm.

[0041] Optionally, A3 includes:

[0042] A31: According to the pre-set conditions of the aneurysm circumscribed rectangular frame, the endpoint combination ABCD is obtained, and the coordinates of the four endpoints A, B, C, and D in the endpoint combination ABCD are calculated; based on the coordinates of the four endpoints A, B, C, and D, the first aneurysm circumscribed rectangular frame is obtained.

[0043] Optionally, it is characterized by:

[0044] The conditions of the pre-set aneurysm circumscribed rectangular frame include:

[0045] The four endpoints in the aneurysm circumscribed rectangular frame are set as A, B, C, and D; the plane formed by the endpoints A, B, and D and the same as the first aneurysm diameter plane is used as the bottom surface of the aneurysm circumscribed rectangular frame, and the opposite side of the bottom surface of the aneurysm circumscribed rectangular frame is used as the top surface of the aneurysm circumscribed rectangular frame;

[0046] The line segment formed by the two endpoints A and B is parallel to the first blood flow direction vector; the line segment formed by the endpoints A and C is used as a side of the aneurysm circumscribed rectangular frame; the top surface of the aneurysm circumscribed rectangular frame passes through the point farthest from the bottom surface of the aneurysm surface point concentration;

[0047] The plane formed by the three endpoints A, C, and D and the opposite side of the plane respectively pass through the two points farthest from each other in the straight line projection formed by the aneurysm surface point set to the endpoints A and B; the plane formed by the three endpoints A, B, and C and the opposite side of the plane respectively pass through the two points farthest from each other in the straight line projection formed by the aneurysm surface point set to the endpoints A and D;

[0048] According to the preset conditions of the aneurysm circumscribed rectangular frame and the aneurysm surface point set, the aneurysm circumscribed rectangular frame is obtained;

[0049] The A31 includes:

[0050] 1) The endpoints of the tumor path are set as N1 and N2 respectively, and the coordinates of N1 are set as (x1, y1, z1), and the coordinates of N2 are set as (x2, y2, z2); the midpoint of the tumor path is set as N C , N C The coordinates are set to (x c ,y c , z c ); set the tumor high point as PH;

[0051] 2) Assume that the plane equation of the tumor diameter is AN*x+BN*y+CN*z=1;

[0052] 3) Solving the first determinant to obtain solution 1, the solution 1 is AF1, BF1, CF1; the plane F1 is represented by the first tumor diameter plane equation AF1*x+BF1*y+CF1*z=1; the first determinant is as follows:

[0053]

[0054] 4) Solving the second determinant to obtain solution 2, the solution 2 is AF2, BF2, CF2; the plane F2 is represented by the second tumor diameter plane equation AF2*x+BF2*y+CF2*z=1; the second determinant is as follows:

[0055]

[0056] 5) The points on the aneurysm surface that are farthest from the two sides of plane F1 are set as pf 11 and pf 12 , the points on the aneurysm surface that are farthest from both sides of plane F2 are set as pf 21 and pf 22 ;

[0057] 6) Set the point pf 11 and pf 12 The projections on plane F1 are set as pf 11 p and pf 12 p, point pf 11 and pf 12 The projections on plane F2 are set as pf 21 p and pf 22 p;

[0058] 7) setting the projection of the aneurysm high point PH on the bottom surface of the aneurysm circumscribed rectangular frame as PHp;

[0059] 8) Based on NC+(pf 12 -pf 12 p)+(pf 21 -pf 21 p), and obtain the coordinates of the endpoint A;

[0060] 9) Based on NC+(pf 11 -pf 11 p)+(pf 21 -pf 21 p), and obtain the coordinates of the endpoint B;

[0061] 10) Based on the vector calculation result of A+(PH-PHp), the coordinates of the endpoint C are obtained;

[0062] 11) Based on NC+(pf 12 -pf 12 p)+(pf 22 -pf 22 p) to obtain the coordinates of the endpoint D.

[0063] Optionally, the A4 includes:

[0064] A41: Based on the coordinates of the four endpoints A, B, C, and D, obtain the coordinates of point E by calculating B+CA, obtain the coordinates of point F by calculating B+DA, obtain the coordinates of point H by calculating C+DA, and obtain the coordinates of point G by calculating C+FA;

[0065] A42: The display interface of the visualization result shows:

[0066] 1) a blood vessel mask in the first mask combination;

[0067] 2) an aneurysm mask in the first mask combination;

[0068] 3) The tumor diameter endpoint, tumor diameter length, tumor diameter midpoint, maximum diameter point, maximum diameter, tumor height point, tumor height projection point, tumor diameter height, tumor width endpoint and tumor width;

[0069] 4) the points A, B, C, D, E, F, G, H, and the line segments AB, BF, AD, DF, AC, BE, DH, FG, CE, EG, CH, HG.

[0070] Optionally, the A5 includes:

[0071] A511: generating an adjustable aneurysm circumscribed rectangular frame on the display interface of the visualization result; adjusting the aneurysm circumscribed rectangular frame accordingly according to the received adjustment instruction to obtain a new aneurysm circumscribed rectangular frame; and using the new aneurysm circumscribed rectangular frame as a second aneurysm circumscribed rectangular frame;

[0072] A512: taking the affine transformation between the second aneurysm circumscribed rectangular frame and the first aneurysm circumscribed rectangular frame as transformation T, and acquiring the transformation T;

[0073] A513: Apply the transformation T to the endpoint combination ABCD to obtain a new endpoint combination A t -B t -C t -D t ; Calculate the new endpoint combination A t -B t -C t -D t Middle A t , B t , C t , D t The coordinates of the four endpoints.

[0074] Optionally, the A5 further includes:

[0075] A521: Based on the pre-set conditions of the aneurysm circumscribed rectangular frame and A t , B t , C t , D t The coordinates of the four endpoints are used to restore the circumscribed rectangular frame of the second aneurysm;

[0076] A522: extracting the portion of the first blood vessel mask within the second aneurysm circumscribed rectangular frame area, marking the connected domains of the extracted first blood vessel mask, removing other connected domains except the largest connected domain in the extracted first blood vessel mask, and obtaining the aneurysm mask in the new mask combination;

[0077] A523: Take the plane where the endpoints At, Bt, and Dt are located as the new tumor diameter plane, and take the geometric parameters of the new tumor diameter plane as the new plane geometric parameters; take the vector formed by the endpoints At and Bt as the new blood flow direction vector; obtain the aneurysm mask in the combination of the new tumor diameter plane, the new blood flow direction vector and the new mask.

[0078] Optionally, the A1 includes:

[0079] For the three-dimensional medical image to be analyzed, segmentation is performed by an automatic segmentation algorithm to obtain a plurality of first aneurysm masks and a plurality of first blood vessel masks; one of the first blood vessel masks matches one of the first aneurysm masks;

[0080] For a first aneurysm mask and an associated first blood vessel mask, performing connected domain division on the first aneurysm mask and the first blood vessel mask to obtain a connected domain within a specified length range to which the first aneurysm mask and the first blood vessel mask belong;

[0081] Extracting a connected domain of a first blood vessel mask and a connected domain of a first aneurysm mask from the three-dimensional medical image, wherein the connected domain of the first blood vessel mask includes the connected domain of the first aneurysm mask;

[0082] removing the connected domain of the first aneurysm mask from the connected domain of the first blood vessel mask to obtain a second blood vessel mask, and converting the connected domain of the first aneurysm mask into a second aneurysm mask;

[0083] An opening operation is performed on the second blood vessel mask through a cubic convolution kernel with a side length of 2 voxels, and the second blood vessel mask of the maximum connected domain after the opening operation and the second aneurysm mask are selected as the first mask combination of a single aneurysm.

[0084] (III) Beneficial effects

[0085] 1) Through machine calculation, the consistency and accuracy of measurement results between doctors can be improved, and the measurement errors caused by different working perspectives between different doctors during manual measurement can be compensated;

[0086] 2) It can reduce the time doctors spend on obtaining aneurysm morphological parameters, improve measurement efficiency, and reduce doctors’ workload;

[0087] 3) The measurement results are more accurate than manual measurement by doctors;

[0088] 4) It can allow doctors to adjust the aneurysm segmentation results to correct the shortcomings caused by the poor performance of the upstream segmentation model;

[0089] 5) The semi-automatic segmentation method of the region of interest can significantly improve the correlation between various morphological parameters and the risk of aneurysm rupture based on accurate and rapid measurement, which is expected to provide auxiliary information for doctors to make treatment decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] The present application is described with the aid of the following drawings:

[0091] Figure 1 The overall process diagram of the semi-automatic measurement method of intracranial aneurysm morphological parameters is shown;

[0092] Figure 2The schematic diagram of the modules involved in the semi-automatic measurement method of intracranial aneurysm morphological parameters is shown;

[0093] Figure 3 It is a schematic diagram of the rectangular frame surrounding the aneurysm;

[0094] Figure 4 It is a schematic diagram of the endpoints and line segments based on the circumscribed rectangular frame of the aneurysm. DETAILED DESCRIPTION

[0095] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below in conjunction with the accompanying drawings through specific implementation methods. It is understood that the specific embodiments described below are only used to explain the relevant inventions, rather than to limit the invention. It should also be noted that the embodiments and features in the embodiments of this application can be combined with each other in the absence of conflict; for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0096] Embodiment 1 provides a semi-automatic measurement method for morphological parameters of intracranial aneurysms. Figure 1 As shown, the first mask combination of a single aneurysm is obtained through a fully automatic preprocessing module; based on the first mask combination of a single aneurysm, the first aneurysm diameter plane and the first blood flow direction vector are obtained through the aneurysm diameter plane and blood flow direction automatic extraction module; based on the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, according to the calculation rules of each parameter value in the parameter information, the parameter information of the aneurysm is obtained through the core calculation module; based on the pre-defined aneurysm circumscribed rectangular frame condition, according to the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, the aneurysm circumscribed rectangular frame is obtained through the circumscribed frame calculation module; based on the first mask combination, the aneurysm circumscribed rectangular frame and the aneurysm parameter information, the result visualization module is used Output visualization results for doctors to view; judge the calculation results based on the output visualization results, and if the calculation results are correct, output parameter information of a single aneurysm; if the calculation results are incorrect, adjust the first aneurysm circumscribed rectangular frame through a manual adjustment module to obtain the endpoint coordinates corresponding to the second aneurysm circumscribed rectangular frame; based on the endpoint coordinates corresponding to the second aneurysm circumscribed rectangular frame, obtain the aneurysm mask in the second aneurysm diameter plane, the second blood flow direction vector and the second mask combination through a semi-automatic preprocessing module; use the aneurysm mask in the second aneurysm diameter plane, the second blood flow direction vector and the second mask combination as the new first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, and recalculate until the calculation results are correct.

[0097] Based on the method described in the first embodiment, the segmentation of the aneurysm and the parent vessel is assisted by manually adjusting the tool, thereby ensuring the accuracy of parameter calculation.

[0098] Embodiment 2 provides a semi-automatic measurement method for morphological parameters of intracranial aneurysms. Figure 2 As shown, the method includes:

[0099] A1. Based on the first mask combination of a single aneurysm, the first aneurysm diameter plane and the first blood flow direction vector are obtained, specifically:

[0100] In this embodiment, a dilation operation is performed on the aneurysm mask in the first mask combination by using a cubic convolution kernel with a side length of 3 voxels, and voxels in the dilation operation result that overlap with the blood vessel mask in the first mask combination are used as intersection points to obtain all intersection points;

[0101] In this embodiment, the geometric center of the intersection point is calculated and used as the estimated center of the tumor body; with the estimated center of the tumor body as the center, a region having the same size as the circumscribed cube of the aneurysm mask in the first mask combination is selected;

[0102] and, taking the coordinate set of all voxels on the aneurysm mask in the first mask combination in the selected region as the aneurysm point set, and taking the coordinate set of all voxels on the blood vessel mask in the first mask combination in the selected region as the blood vessel point set; obtaining the aneurysm point set and the blood vessel point set;

[0103] In this embodiment, a support vector machine classification model is established based on the aneurysm point set and the blood vessel point set;

[0104] And, fitting through the support vector machine classification model to obtain a classification plane; using the obtained classification plane as the first tumor diameter plane, and using the geometric parameters of the first tumor diameter plane as the first plane geometric parameters; obtaining the first tumor diameter plane and the first plane geometric parameters;

[0105] In this embodiment, a cubic region with a side length of 15 voxels is selected with the estimated center of the tumor as the center; the blood vessel mask in the first mask combination in the selected cubic region is subjected to three-dimensional skeletonization processing to obtain a skeleton point set; if a point in the skeleton point set satisfies the condition that there is only one point belonging to the skeleton point set in the 26 neighborhoods of the point, the point is used as the starting point; and all starting points are obtained;

[0106] And, randomly select a point from the acquired starting points as the first starting point, set the first starting point as the current point, and set the point closest to the current point and not yet the current point as the new current point; search for the current point in the skeleton point set until the current point is another starting point;

[0107] And, another starting point is used as the second starting point; the projection of the vector formed by the first starting point and the second starting point on the first tumor diameter plane is used as the first blood flow direction vector to obtain the first blood flow direction vector.

[0108] A2. Based on the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, according to the calculation rules of each parameter value in the parameter information, the parameter information of the aneurysm is obtained, wherein the parameter information of the aneurysm includes one or more of the following: aneurysm diameter length, maximum diameter, aneurysm height and aneurysm width, specifically:

[0109] In this embodiment, based on the first plane geometric parameters, the distance between each voxel on the aneurysm mask in the first mask combination and the first aneurysm diameter plane is calculated, and the set of points whose distance is less than one voxel unit is taken as the first aneurysm diameter point set; the first aneurysm diameter point set is obtained;

[0110] In this embodiment, all points on the aneurysm mask in the first mask combination are traversed. If there is a point on the aneurysm mask that does not belong to the first mask combination within the 18 neighborhood of the current traversed point, the current traversed point is taken as the aneurysm surface point, and the set of all aneurysm surface points is taken as the aneurysm surface point set; the aneurysm surface point set is obtained;

[0111] In this embodiment, the point set combinations of any two points in the first tumor diameter point set are traversed; the Euclidean distance between the two points in all point set combinations is calculated, and the absolute value of the cosine of the angle between the vector formed by the two points in all point set combinations and the first blood flow direction vector is calculated; the point set combinations whose absolute value of the cosine of the angle is less than 0.9 are removed to obtain the second tumor diameter point set;

[0112] And, based on the second tumor path point set, the maximum Euclidean distance between two points in the point set combination is screened out, and the maximum Euclidean distance is used as the tumor path length; the two points corresponding to the tumor path length are used as the tumor path endpoints, and the midpoint of the tumor path endpoints is used as the tumor path midpoint; the tumor path length, the tumor path endpoints and the tumor path midpoint are obtained;

[0113] In this embodiment, the Euclidean distances between all points in the aneurysm surface point set and the midpoint of the aneurysm diameter are calculated, the point farthest from the midpoint of the aneurysm diameter is taken as the maximum diameter point, and the Euclidean distance between the maximum diameter point and the midpoint of the aneurysm diameter is taken as the maximum diameter; the vector formed by the maximum diameter point and the midpoint of the aneurysm diameter is taken as the maximum diameter direction vector; the maximum diameter point, the maximum diameter and the maximum diameter direction vector are obtained;

[0114] In this embodiment, for the points in the aneurysm surface point concentration, the points that are not on the same side of the aneurysm diameter plane as the maximum diameter point are deleted;

[0115] In this embodiment, the Euclidean distances between all points in the aneurysm surface point set and the first aneurysm diameter plane are calculated, the point farthest from the first aneurysm diameter plane is taken as the aneurysm height point, and the Euclidean distance between the aneurysm height point and the first aneurysm diameter plane is taken as the aneurysm body height; the projection point of the aneurysm height point on the first aneurysm diameter plane is taken as the aneurysm height projection point, and the projection point is taken as the aneurysm height projection point; the aneurysm height point, the aneurysm body height and the aneurysm height projection point are obtained;

[0116] In this embodiment, the point set combinations of any two points in the aneurysm surface point set are traversed; the Euclidean distance between the two points in all point set combinations is calculated, and the absolute value of the cosine of the angle between the vector formed by the two points in all point set combinations and the first blood flow direction vector is calculated; the point set combinations whose absolute value of the cosine of the angle is less than 0.3 are removed to obtain the third aneurysm diameter point set;

[0117] And, for the third tumor diameter point set, select the maximum Euclidean distance between two points in the point set combination, use the maximum Euclidean distance as the tumor diameter width, and use the two points corresponding to the tumor diameter width as the tumor width endpoints; obtain the tumor diameter width and the tumor width endpoints;

[0118] In this embodiment, the aneurysm diameter length, maximum diameter, aneurysm height and aneurysm width are used as parameter information of a single aneurysm, and the parameter information of a single aneurysm is merged into a parameter information list of a single aneurysm; and the parameter information list of a single aneurysm is obtained.

[0119] A3. Based on the predefined condition of the aneurysm circumscribed rectangular frame, the aneurysm circumscribed rectangular frame is obtained according to the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, specifically:

[0120] In this embodiment, Figure 3 As shown, according to the pre-set conditions of the aneurysm circumscribed rectangular frame, the endpoint combination ABCD is obtained, and the coordinates of the four endpoints A, B, C, and D in the endpoint combination ABCD are calculated; based on the coordinates of the four endpoints A, B, C, and D, the first aneurysm circumscribed rectangular frame is obtained.

[0121] In this embodiment, the conditions of the pre-set aneurysm circumscribed rectangular frame include: setting the four endpoints in the aneurysm circumscribed rectangular frame as A, B, C, and D; setting the plane formed by the endpoints A, B, and D, which is the same as the first aneurysm diameter plane, as the bottom surface of the aneurysm circumscribed rectangular frame, and setting the opposite side of the bottom surface of the aneurysm circumscribed rectangular frame as the top surface of the aneurysm circumscribed rectangular frame;

[0122] And, the line segment formed by the two endpoints A and B is parallel to the first blood flow direction vector; the line segment formed by the endpoints A and C is used as an edge of the aneurysm circumscribed rectangular frame; the top surface of the aneurysm circumscribed rectangular frame passes through the point farthest from the bottom surface of the aneurysm surface point concentration;

[0123] Furthermore, the plane formed by the three endpoints A, C, and D and the opposite side of the plane pass through the two points farthest from each other in the projection of the line formed by the aneurysm surface point set to the endpoints A and B; the plane formed by the three endpoints A, B, and C and the opposite side of the plane pass through the two points farthest from each other in the projection of the line formed by the aneurysm surface point set to the endpoints A and D;

[0124] and, obtaining an aneurysm circumscribed rectangular frame according to a preset condition of the aneurysm circumscribed rectangular frame and a set of aneurysm surface points;

[0125] In addition, in this embodiment, according to the definition of the aneurysm circumscribed rectangular frame set in advance, the endpoint combination ABCD is obtained, and the coordinates of the four endpoints A, B, C, and D in the endpoint combination ABCD are calculated as follows:

[0126] 1) Set the endpoints of the tumor path to N1 and N2, and set the coordinates of N1 to (x1, y1, z1), and the coordinates of N2 to (x2, y2, z2); set the midpoint of the tumor path to N C , N C The coordinates are set to (x c ,y c , z c ); set the tumor high point as PH;

[0127] 2) Assume that the plane equation of the tumor diameter is AN*x+BN*y+CN*z=1;

[0128] 3) Solution 1 is obtained by solving the first determinant, and solution 1 is AF1, BF1, CF1; plane F1 is represented by the first tumor diameter plane equation AF1*x+BF1*y+CF1*z=1; the first determinant is as follows:

[0129]

[0130] 4) Solving the second determinant to obtain solution 2, solution 2 is AF2, BF2, CF2; plane F2 is represented by the second tumor diameter plane equation AF2*x+BF2*y+CF2*z=1; the second determinant is as follows:

[0131]

[0132] 5) Set the points on the aneurysm surface that are farthest from the two sides of plane F1 as pf 11 and pf 12 , the points on the aneurysm surface that are farthest from the two sides of plane F2 are set as pf 21 and pf 22 ;

[0133] 6) Set the point pf 11 and pf 12 The projections on plane F1 are set as pf 11 p and pf 12 p, point pf 11 and pf 12 The projections on plane F2 are set as pf 21 p and pf 22p;

[0134] 7) The projection of the aneurysm high point PH on the bottom surface of the aneurysm circumscribed rectangular frame is set as PHp;

[0135] 8) Based on NC+(pf 12 -pf 12 p)+(pf 21 -pf 21 p) to obtain the coordinates of endpoint A;

[0136] 9) Based on NC+(pf 11 -pf 11 p)+(pf 21 -pf 21 p) to obtain the coordinates of endpoint B;

[0137] 10) Based on the vector calculation result of A+(PH-PHp), obtain the coordinates of the endpoint C;

[0138] 11) Based on NC+(pf 12 -pf 12 p)+(pf 22 -pf 22 p) to obtain the coordinates of the endpoint D.

[0139] A4. Based on the first mask combination, the aneurysm circumscribed rectangular frame and the aneurysm parameter information, output a visualization result for the doctor to view, specifically:

[0140] In this embodiment, based on the coordinates of the four endpoints A, B, C, and D, the coordinates of point E are obtained by the calculation method of B+CA, the coordinates of point F are obtained by the calculation method of B+DA, the coordinates of point H are obtained by the calculation method of C+DA, and the coordinates of point G are obtained by the calculation method of C+FA;

[0141] In this embodiment, the display interface of the visualization result shows:

[0142] 1) Vessel mask in the first mask combination;

[0143] 2) the aneurysm mask in the first mask combination;

[0144] 3) Tumor diameter endpoint, tumor diameter length, tumor diameter midpoint, maximum diameter point, maximum diameter, tumor height point, tumor height projection point, tumor diameter height, tumor width endpoint and tumor width;

[0145] 4) Points A, B, C, D, E, F, G, H, and line segments AB, BF, AD, DF, AC, BE, DH, FG, CE, EG, CH, HG Figure 4 shown.

[0146] Based on the method described in Example 2, firstly, through machine calculation, the consistency and accuracy of the measurement results between doctors can be improved, and the measurement errors caused by different doctors choosing different working perspectives during manual measurement can be compensated; secondly, the time for doctors to obtain aneurysm morphological parameters can be reduced, the measurement efficiency can be improved, and the workload of doctors can be reduced.

[0147] Embodiment 3 provides a method for semi-automatic measurement of morphological parameters of intracranial aneurysms, the method comprising:

[0148] The first mask combination is information about the connected domain of a single aneurysm mask and the connected domain of a blood vessel mask of a blood vessel to which the single aneurysm mask belongs, which are pre-extracted from the three-dimensional medical image to be analyzed, and is specifically:

[0149] In this embodiment, the three-dimensional medical image to be analyzed is segmented by an automatic segmentation algorithm to obtain a plurality of first aneurysm masks and a plurality of first blood vessel masks; a first blood vessel mask matches a first aneurysm mask;

[0150] In this embodiment, for a first aneurysm mask and an associated first blood vessel mask, the first aneurysm mask and the first blood vessel mask are divided into connected domains to obtain connected domains within a specified length range to which the first aneurysm mask and the first blood vessel mask belong;

[0151] Specifically, the number of single connected domains corresponds to the number of single aneurysm masks;

[0152] Specifically, taking the centroid of a single connected domain as the center, a cubic region with 64 voxels in length, width and height is obtained; a blood vessel mask of the cubic region and an aneurysm mask corresponding to the single connected domain are obtained;

[0153] In this embodiment, a connected domain of a first blood vessel mask and a connected domain of a first aneurysm mask are extracted from the three-dimensional medical image, and the connected domain of the first blood vessel mask includes a connected domain of the first aneurysm mask;

[0154] Specifically, the connected domain of the first blood vessel mask is marked, and if there is an overlapping portion between the first blood vessel mask and the first aneurysm mask, the connected domain to which the overlapping portion belongs is retained based on the connected domain of the first blood vessel mask;

[0155] In this embodiment, the connected domain of the first aneurysm mask is removed from the connected domain of the first blood vessel mask to obtain a second blood vessel mask, and the connected domain of the first aneurysm mask is converted into a second aneurysm mask;

[0156] Specifically, the overlapping portion of the first blood vessel mask and the first aneurysm mask is removed to obtain a second blood vessel mask;

[0157] In this embodiment, an opening operation is performed on the second blood vessel mask by using a cubic convolution kernel with a side length of 2 voxels, and the second blood vessel mask and the second aneurysm mask of the maximum connected domain after the opening operation are selected as the first mask combination of a single aneurysm;

[0158] Specifically, the second blood vessel mask is processed by an opening operation, so that abnormal concave and convex points of the blood vessel image can be scattered;

[0159] Specifically, other connected domains other than the largest connected domain after the opening operation are removed, that is, the tiny connected domains formed by the scattered abnormal concave and convex points are removed, so that the blood vessel image is smoother and more accurate parameters are obtained.

[0160] Based on the method described in the third embodiment, the first mask combination of a single aneurysm is obtained, so that targeted parameter measurement of the single aneurysm can be achieved; in addition, the vascular mask is processed by an open operation, so that the parameter calculation result can be made more accurate.

[0161] Embodiment 4 provides a semi-automatic measurement method for morphological parameters of intracranial aneurysms, specifically:

[0162] In this embodiment, the calculation result is judged according to the output visualization result, and if the calculation result is correct, the parameter information of the single aneurysm is output;

[0163] Furthermore, if the calculation result is incorrect, the three-dimensional medical image is re-segmented based on the adjustment instructions received by the display interface of the visualization result to obtain the aneurysm mask in the second aneurysm diameter plane, the second blood flow direction vector and the second mask combination; based on the aneurysm mask in the second aneurysm diameter plane, the second blood flow direction vector and the second mask combination, the process of steps A2 to A5 is repeated until the calculation result is correct.

[0164] In this embodiment, the first mask combination corresponding to one or more connected domains is obtained through a fully automatic preprocessing module. The specific process has been described above and will not be repeated here.

[0165] In this embodiment, if the number of connected domains is 1, the corresponding parameter information of a single aneurysm is output based on the first mask combination obtained; the aneurysm morphological parameter is the output parameter information of the single aneurysm;

[0166] In this embodiment, if the number of connected domains is greater than 1, the parameter information of all output single aneurysms is merged into the parameter information of multiple aneurysms; the aneurysm morphological parameters are the parameter information of the output multiple aneurysms;

[0167] In this embodiment, the first aneurysm circumscribed rectangular frame is adjusted by a manual adjustment module to obtain the endpoint coordinates corresponding to the second aneurysm circumscribed rectangular frame;

[0168] In this embodiment, based on the endpoint coordinates corresponding to the circumscribed rectangular frame of the second aneurysm, the aneurysm mask in the second aneurysm diameter plane, the second blood flow direction vector and the second mask combination is obtained through a semi-automatic preprocessing module.

[0169] Based on the method described in the fourth embodiment, doctors can adjust the aneurysm segmentation results, thereby correcting the shortcomings caused by the poor effect of the upstream segmentation model.

[0170] Embodiment 5 provides a semi-automatic measurement method for morphological parameters of intracranial aneurysms, specifically:

[0171] In this embodiment, an adjustable aneurysm circumscribed rectangular frame is generated on the display interface of the visualization result; the aneurysm circumscribed rectangular frame is adjusted accordingly according to the received adjustment instruction to obtain a new aneurysm circumscribed rectangular frame; the new aneurysm circumscribed rectangular frame is used as the second aneurysm circumscribed rectangular frame;

[0172] In this embodiment, the affine transformation between the second aneurysm circumscribed rectangular frame and the first aneurysm circumscribed rectangular frame is used as the transformation T to obtain the transformation T;

[0173] In this embodiment, the transformation T is applied to the endpoint combination ABCD to obtain a new endpoint combination A t -B t -C t -D t ; Calculate the new endpoint combination A t -B t -C t -D t Middle A t , B t , C t , D t The coordinates of the four endpoints.

[0174] Based on the method described in the fifth embodiment, by adjusting the aneurysm circumscribed rectangular frame, a technical basis is provided for doctors to adjust the aneurysm segmentation result.

[0175] Embodiment 6 provides a semi-automatic measurement method for morphological parameters of intracranial aneurysms, specifically:

[0176] In this embodiment, according to the pre-set conditions of the aneurysm circumscribed rectangular frame and A t , B t , C t , D t The coordinates of the four endpoints are used to restore the circumscribed rectangular frame of the second aneurysm;

[0177] In this embodiment, a portion of the first blood vessel mask within the second aneurysm circumscribed rectangular frame region is extracted, connected domains are marked on the extracted first blood vessel mask, and other connected domains other than the largest connected domain in the extracted first blood vessel mask are removed to obtain an aneurysm mask in a new mask combination;

[0178] In this embodiment, the plane where the endpoints At, Bt, and Dt are located is used as the new tumor diameter plane, and the geometric parameters of the new tumor diameter plane are used as the new plane geometric parameters; the vector formed by the endpoints At and Bt is used as the new blood flow direction vector; and the aneurysm mask in the new tumor diameter plane, the new blood flow direction vector, and the new mask combination is obtained.

[0179] Based on the method described in Example 6, the doctor can select a portion of interest in the three-dimensional medical image and measure the parameters of the selected portion, which is more flexible and convenient to operate.

[0180] Embodiment 7 provides a semi-automatic measurement method for morphological parameters of intracranial aneurysms, which classifies the vascular regions of interest by means of skeleton topology analysis, and uses the GAR method to find the opening of the aneurysm in the region of interest, thereby segmenting the aneurysm and measuring the height and width of the aneurysm.

[0181] Based on the method described in the seventh embodiment, the doctor can select a portion of interest in the three-dimensional medical image and measure the parameters of the selected portion, which is more flexible and convenient to operate.

[0182] Embodiment 8 provides a semi-automatic measurement method for the morphological parameters of intracranial aneurysms. The semi-automatic method is used to continuously place seed spheres in the blood vessels with aneurysms, determine the different densities of the blood vessels and surrounding tissues, extract and segment the aneurysm sac in combination with the dome calculation method, automatically draw the aneurysm neck line, and achieve further measurement.

[0183] Based on the method described in Example 8, the doctor can select a portion of interest in the three-dimensional medical image and measure the parameters of the selected portion, which is more flexible and convenient to operate.

[0184] This embodiment also provides an electronic device, including: a memory and a processor; the processor is used to execute the computer program stored in the memory to implement the steps of the semi-automatic measurement method of intracranial aneurysm morphological parameters described in any of the above embodiments. The electronic device can be a computer, or a device belonging to a client commonly used by doctors.

[0185] Specifically, the electronic device of this embodiment may include: at least one processor, at least one memory, at least one network interface and / or other user interface. The various components in the electronic device are coupled together through a bus system. It can be understood that the bus system is used to realize the connection and communication between these components. In addition to the data bus, the bus system also includes a power bus, a control bus and a status signal bus.

[0186] The electronic device of this embodiment can execute Figures 1 to 4 In any of the methods shown, the user interface may include a display, a keyboard, or a pointing device (eg, a mouse or a touch pad, etc.).

[0187] It can be understood that the memory in this embodiment can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. The memory described herein includes any other suitable type of memory.

[0188] In some embodiments, the memory stores the following elements, executable units or data structures, or a subset thereof, or an extended set thereof: an operating system and application programs.

[0189] The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic services and process hardware-based tasks. The application includes various application programs, such as a media player (MediaPlayer), a browser (Browser), etc., which are used to implement various application services. The program for implementing the method of the embodiment of the present invention can be included in the application.

[0190] In an embodiment of the present invention, the processor calls a program or instruction stored in a memory, specifically, a program or instruction stored in an application program, and the processor is used to execute the method steps provided in the first aspect.

[0191] The method disclosed in the above embodiment of the present invention can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiment of the present invention can be directly embodied as a hardware decoding processor to be executed, or a combination of hardware and software units in the decoding processor to be executed. The software unit may be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0192] On the other hand, an embodiment of the present invention further provides a computer-readable storage medium, which is used to store a computer program, and when the computer program is executed by a processor, the steps of the semi-automatic measurement method of any of the above embodiments are implemented.

[0193] In particular, the semi-automatic measurement method uses some conditions or information pre-defined by the user or triggered by the user, which are pre-entered through the user interface. The main implementation is a technical solution with more accurate measurement results and simpler measurement methods.

[0194] It should be noted that in the claims, any figure marks between brackets should not be understood as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "one" or "an" preceding a component does not exclude the presence of multiple such components. In addition, it should be noted that in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" and the like refers to the specific features, structures, materials or characteristics described in conjunction with the embodiment or example included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0195] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments after knowing the basic creative concept. Therefore, the claims should be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0196] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention should also include these modifications and variations.

Claims

1. A semi-automatic measurement method for morphological parameters of intracranial aneurysms, characterized in that: include: A1. Based on the first mask combination of a single aneurysm, a first aneurysm diameter plane and a first blood flow direction vector are obtained; The first mask combination is information of a connected domain of a single aneurysm mask and information of a connected domain of a blood vessel mask of a blood vessel to which the single aneurysm mask belongs, which are pre-extracted from the three-dimensional medical image to be analyzed; A2. based on the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, and according to the calculation rule of each parameter value in the parameter information, obtaining parameter information of the aneurysm; The parameter information of the aneurysm includes one or more of the following: aneurysm diameter length, maximum diameter, aneurysm height and aneurysm width; A3, based on the predefined condition of the aneurysm circumscribed rectangular frame, according to the first aneurysm diameter plane, the first blood flow direction vector and the aneurysm mask in the first mask combination, obtaining the aneurysm circumscribed rectangular frame; The A3 includes: A31, according to the pre-set conditions of the aneurysm circumscribed rectangular frame, obtaining the endpoint combination ABCD, calculating the coordinates of the four endpoints A, B, C, and D in the endpoint combination ABCD; based on the coordinates of the four endpoints A, B, C, and D, obtaining the first aneurysm circumscribed rectangular frame; The conditions of the pre-set aneurysm circumscribed rectangular frame include: The four endpoints in the aneurysm circumscribed rectangular frame are set as A, B, C, and D; the plane formed by the endpoints A, B, and D and the same as the first aneurysm diameter plane is used as the bottom surface of the aneurysm circumscribed rectangular frame, and the opposite side of the bottom surface of the aneurysm circumscribed rectangular frame is used as the top surface of the aneurysm circumscribed rectangular frame; The line segment formed by the two endpoints A and B is parallel to the first blood flow direction vector; the line segment formed by the endpoints A and C is used as a side of the aneurysm circumscribed rectangular frame; the top surface of the aneurysm circumscribed rectangular frame passes through the point farthest from the bottom surface of the aneurysm surface point concentration; The plane formed by the three endpoints A, C, and D and the opposite side of the plane respectively pass through the two points farthest from each other in the straight line projection formed by the aneurysm surface point set to the endpoints A and B; the plane formed by the three endpoints A, B, and C and the opposite side of the plane respectively pass through the two points farthest from each other in the straight line projection formed by the aneurysm surface point set to the endpoints A and D; According to the preset conditions of the aneurysm circumscribed rectangular frame and the aneurysm surface point set, the aneurysm circumscribed rectangular frame is obtained; A4. Based on the first mask combination, the aneurysm circumscribed rectangular frame and the parameter information of the aneurysm, a visualization result is output for the doctor to view.

2. The method according to claim 1, characterized in that The method further comprises: A5. If the display interface of the visualization result receives an adjustment instruction triggered by the doctor; Then, according to the adjustment parameters in the adjustment instruction, based on re-segmenting the three-dimensional medical image, a new aneurysm diameter plane, a new blood flow direction vector and an aneurysm mask in a new mask combination are obtained; the new aneurysm diameter plane is used as a second aneurysm diameter plane; the new blood flow direction vector is used as a second blood flow direction vector; and the aneurysm mask in the new mask combination is used as an aneurysm mask in the second mask combination; For the second aneurysm diameter plane, the second blood flow direction vector and the aneurysm mask in the second mask combination corresponding to a single aneurysm, the second aneurysm diameter plane is used as the new first aneurysm diameter plane, the second blood flow direction vector is used as the new first blood flow direction vector, and the aneurysm mask in the second mask combination is used as the aneurysm mask in the new first mask combination; repeat the above processes A2 to A5.

3. The method according to claim 1, characterized in that The A1 includes: A11: performing a dilation operation on the aneurysm mask in the first mask combination by using a cubic convolution kernel with a side length of 3 voxels, taking voxels in the dilation operation result that overlap with the blood vessel mask in the first mask combination as intersection points, and obtaining all intersection points; A12: Calculate the geometric center of the intersection point, and use the geometric center as the estimated center of the tumor body; and select a region with the same size as the circumscribed cube of the aneurysm mask in the first mask combination, with the estimated center of the tumor body as the center; Taking the coordinate set of all voxels on the aneurysm mask in the first mask combination in the selected area as the aneurysm point set, and taking the coordinate set of all voxels on the blood vessel mask in the first mask combination in the selected area as the blood vessel point set; acquiring the aneurysm point set and the blood vessel point set; A13: establishing a support vector machine classification model based on the aneurysm point set and the blood vessel point set; Fitting the support vector machine classification model to obtain a classification plane; using the obtained classification plane as a first tumor diameter plane, and using the geometric parameters of the first tumor diameter plane as first plane geometric parameters; obtaining the first tumor diameter plane and the first plane geometric parameters; A14: taking the estimated center of the tumor as the center, select a cubic region with a side length of 15 voxels; perform three-dimensional skeletonization on the blood vessel mask in the first mask combination in the selected cubic region to obtain a skeleton point set; if a point in the skeleton point set satisfies the condition that there is only one point belonging to the skeleton point set in the 26 neighborhoods of the point, then take the point as the starting point; obtain all starting points; A point is randomly selected from the acquired starting points as a first starting point, the first starting point is set as a current point, and a point which is closest to the current point and has not been the current point is set as a new current point; the current point is searched in the skeleton point set until the current point is another starting point; The other starting point is used as the second starting point; the projection of the vector formed by the first starting point and the second starting point on the first tumor diameter plane is used as the first blood flow direction vector to obtain the first blood flow direction vector.

4. The method according to claim 3, characterized in that The A2 includes: A21: based on the first plane geometric parameters, calculate the distance between each voxel on the aneurysm mask in the first mask combination and the first aneurysm diameter plane, and take the set of points whose distance is less than one voxel unit as the first aneurysm diameter point set; obtain the first aneurysm diameter point set; A22: traverse all points on the aneurysm mask in the first mask combination, and if there is a point on the aneurysm mask that does not belong to the first mask combination within the 18 neighborhood of the current traversal point, take the current traversal point as an aneurysm surface point, and take the set of all aneurysm surface points as an aneurysm surface point set; obtain the aneurysm surface point set; A23: traversing any two point combinations in the first tumor diameter point set; calculating the Euclidean distance between the two points in all point set combinations, and calculating the absolute value of the cosine of the angle between the vector formed by the two points in all point set combinations and the first blood flow direction vector; removing the point set combinations whose absolute value of the cosine of the angle is less than 0.9, to obtain the second tumor diameter point set; Based on the second tumor path point set, the maximum Euclidean distance between two points in the point set combination is screened out, and the maximum Euclidean distance is used as the tumor path length; the two points corresponding to the tumor path length are used as the tumor path endpoints, and the midpoint of the tumor path endpoints is used as the tumor path midpoint; and the tumor path length, the tumor path endpoints and the tumor path midpoint are obtained; A24: Calculate the Euclidean distances between all points in the aneurysm surface point set and the midpoint of the aneurysm diameter, take the point farthest from the midpoint of the aneurysm diameter as the maximum diameter point, take the Euclidean distance between the maximum diameter point and the midpoint of the aneurysm diameter as the maximum diameter; take the vector formed by the maximum diameter point and the midpoint of the aneurysm diameter as the maximum diameter direction vector; obtain the maximum diameter point, maximum diameter and maximum diameter direction vector; A25: for the points in the aneurysm surface point set, delete the points that are not on the same side of the aneurysm diameter plane as the maximum diameter point; A26: Calculate the Euclidean distance between all points in the aneurysm surface point set and the first aneurysm diameter plane, take the point farthest from the first aneurysm diameter plane as the aneurysm height point, take the Euclidean distance between the aneurysm height point and the first aneurysm diameter plane as the aneurysm body height; take the projection point of the aneurysm height point on the first aneurysm diameter plane as the aneurysm height projection point, and take the projection point as the aneurysm height projection point; obtain the aneurysm height point, aneurysm body height and aneurysm height projection point; A27: traverse the point set combination of any two points in the aneurysm surface point set; calculate the Euclidean distance between the two points in all point set combinations, calculate the absolute value of the cosine of the angle between the vector formed by the two points in all point set combinations and the first blood flow direction vector; remove the point set combination whose absolute value of the cosine of the angle is less than 0.3, and obtain the third aneurysm diameter point set; For the third tumor diameter point set, select the maximum Euclidean distance between two points in the point set combination, use the maximum Euclidean distance as the tumor diameter width, and use the two points corresponding to the tumor diameter width as the tumor width endpoints; obtain the tumor diameter width and the tumor width endpoints; A28: taking the aneurysm diameter length, maximum diameter, aneurysm height and aneurysm width as parameter information of a single aneurysm, and merging the parameter information of the single aneurysm into a parameter information list of the single aneurysm; and obtaining the parameter information list of the single aneurysm.

5. The method according to claim 4, characterized in that: The A31 includes: 1) The endpoints of the tumor path are set as N1 and N2 respectively, and the coordinates of N1 are set as (x1, y1, z1), and the coordinates of N2 are set as (x2, y2, z2); the midpoint of the tumor path is set as N C , N C The coordinates are set to (x c ,y c , z c ); set the tumor high point as PH; 2) Assume that the plane equation of the tumor diameter is AN*x+BN*y+CN*z=1; 3) Solving the first determinant to obtain solution 1, the solution 1 is AF1, BF1, CF1; the plane F1 is represented by the first tumor diameter plane equation AF1*x+BF1*y+CF1*z=1; the first determinant is as follows: 4) Solving the second determinant to obtain solution 2, the solution 2 is AF2, BF2, CF2; the plane F2 is represented by the second tumor diameter plane equation AF2*x+BF2*y+CF2*z=1; the second determinant is as follows: 5) The points on the aneurysm surface that are farthest from the two sides of plane F1 are set as pf 11 and pf 12 , the points on the aneurysm surface that are farthest from both sides of plane F2 are set as pf 21 and pf 22 ; 6) Set the point pf 11 and pf 12 The projections on plane F1 are set as pf 11 p and pf 12 p, point pf 11 and pf 12 The projections on plane F2 are set as pf 21 p and pf 22 p; 7) The projection of the aneurysm high point PH on the bottom surface of the aneurysm circumscribed rectangular frame is set as PHp; 8) Based on NC+(pf 12 -pf 12 p)+(pf 21 -pf 21 p), and obtain the coordinates of the endpoint A; 9) Based on NC+(pf 11 -pf 11 p)+(pf 21 -pf 21 p), and obtain the coordinates of the endpoint B; 10) Based on the vector calculation result of A+(PH-PHp), the coordinates of the endpoint C are obtained; 11) Based on NC+(pf 12 -pf 12 p)+(pf 22 -pf 22 p) to obtain the coordinates of the endpoint D.

6. The method according to claim 5, characterized in that The A4 includes: A41: Based on the coordinates of the four endpoints A, B, C, and D, obtain the coordinates of point E by calculating B+CA, obtain the coordinates of point F by calculating B+DA, obtain the coordinates of point H by calculating C+DA, and obtain the coordinates of point G by calculating C+FA; A42: The display interface of the visualization result shows: 1) a blood vessel mask in the first mask combination; 2) an aneurysm mask in the first mask combination; 3) The tumor diameter endpoint, tumor diameter length, tumor diameter midpoint, maximum diameter point, maximum diameter, tumor height point, tumor height projection point, tumor diameter height, tumor width endpoint and tumor width; 4) the points A, B, C, D, E, F, G, H, and the line segments AB, BF, AD, DF, AC, BE, DH, FG, CE, EG, CH, HG.

7. The method according to claim 2, characterized in that: The A5 includes: A511: generating an adjustable aneurysm circumscribed rectangular frame on the display interface of the visualization result; adjusting the aneurysm circumscribed rectangular frame accordingly according to the received adjustment instruction to obtain a new aneurysm circumscribed rectangular frame; and using the new aneurysm circumscribed rectangular frame as a second aneurysm circumscribed rectangular frame; A512: taking the affine transformation between the second aneurysm circumscribed rectangular frame and the first aneurysm circumscribed rectangular frame as transformation T, and acquiring the transformation T; A513: Apply the transformation T to the endpoint combination ABCD to obtain a new endpoint combination A t -B t -C t -D t ; Calculate the new endpoint combination A t -B t -C t -D t Middle A t , B t , C t , D t The coordinates of the four endpoints.

8. The method according to claim 7, characterized in that The A5 also includes: A521: Based on the pre-set conditions of the aneurysm circumscribed rectangular frame and A t , B t , C t , D t The coordinates of the four endpoints are used to restore the circumscribed rectangular frame of the second aneurysm; A522: extracting the portion of the first blood vessel mask within the second aneurysm circumscribed rectangular frame area, marking the connected domains of the extracted first blood vessel mask, removing other connected domains except the largest connected domain in the extracted first blood vessel mask, and obtaining the aneurysm mask in the new mask combination; the first blood vessel mask is the blood vessel mask in the first mask combination; A523: Endpoint A t , B t , D t The plane where the end point A is located is taken as the new tumor diameter plane, and the geometric parameters of the new tumor diameter plane are taken as the new plane geometric parameters; t , B t The constructed vector is used as a new blood flow direction vector; and an aneurysm mask in the combination of the new aneurysm diameter plane, the new blood flow direction vector and the new mask is obtained.

9. The method according to claim 1, characterized in that: The A1 previously includes: For the three-dimensional medical image to be analyzed, segmentation is performed by an automatic segmentation algorithm to obtain a plurality of first aneurysm masks and a plurality of first blood vessel masks; one of the first blood vessel masks matches one of the first aneurysm masks; For a first aneurysm mask and an associated first blood vessel mask, performing connected domain division on the first aneurysm mask and the first blood vessel mask to obtain a connected domain within a specified length range to which the first aneurysm mask and the first blood vessel mask belong; Extracting a connected domain of a first blood vessel mask and a connected domain of a first aneurysm mask from the three-dimensional medical image, wherein the connected domain of the first blood vessel mask includes the connected domain of the first aneurysm mask; removing the connected domain of the first aneurysm mask from the connected domain of the first blood vessel mask to obtain a second blood vessel mask, and converting the connected domain of the first aneurysm mask into a second aneurysm mask; An opening operation is performed on the second blood vessel mask through a cubic convolution kernel with a side length of 2 voxels, and the second blood vessel mask of the maximum connected domain after the opening operation and the second aneurysm mask are selected as the first mask combination of a single aneurysm.

Images