Intracranial aneurysm localization method and device based on digitization

Abstract

The present invention provides a method and device for locating intracranial aneurysms based on digitization. The method includes: obtaining a patient's brain image, identifying it, and determining each aneurysm tissue therein; performing image interception on the brain image based on each aneurysm tissue to obtain each tumor tissue area of ​​each aneurysm tissue; establishing each image processing layer corresponding to each tumor tissue area, and filling each tumor tissue area into each image processing layer; determining a predicted burst point based on the tissue morphology of each aneurysm tissue, and generating a rupture prediction area in the image processing layer based on the predicted burst point; performing image combination of each tumor tissue area and the corresponding rupture prediction area to obtain each initial prediction image in each image processing layer; performing content correction on the blank area located in each initial prediction image based on the brain image to obtain each current prediction image. The present invention at least improves the accuracy of predicting the range of aneurysm rupture.

Description

Technical Field

[0001] The invention relates to data processing technology, and in particular to a digital-based intracranial aneurysm positioning method and device. Background Art

[0002] Intracranial aneurysms are mostly abnormal bulges on the walls of intracranial arteries. They are the leading cause of subarachnoid hemorrhage and the third most common cause of cerebrovascular accidents, after cerebral thrombosis and hypertensive cerebral hemorrhage. They can occur at any age, but are more common in middle-aged and elderly women between 40 and 60 years old. For the treatment of intracranial aneurysms, a clear diagnosis must first be made through cerebral angiography, followed by early surgery and interventional embolization, because once an intracranial aneurysm ruptures, the mortality rate is extremely high.

[0003] The inventors found in their research that the existing method for locating intracranial aneurysms is usually performed by doctors by checking the patient's cerebral angiography, which lacks digital equipment to assist and has certain limitations. Summary of the invention

[0004] Based on the above problems, the present invention is proposed to provide a digital-based intracranial aneurysm positioning method and device that overcomes the above problems or at least partially solves the above problems.

[0005] According to one aspect of the present invention, a digital-based intracranial aneurysm positioning method is provided, comprising the following steps:

[0006] Acquiring a brain image of the patient, and identifying the brain image to determine each aneurysmal tissue present in the brain image;

[0007] Performing regional interception on the brain image based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue;

[0008] Establishing image processing layers corresponding to the tumor tissue regions, and filling the tumor tissue regions into the corresponding image processing layers;

[0009] Determining corresponding predicted bursting points based on the tissue morphology of each aneurysm tissue, and generating a rupture prediction area in the image processing layer based on the predicted bursting points;

[0010] Combining images of each tumor tissue region with the corresponding rupture prediction region to obtain initial prediction images with rectangular outlines located in each of the image processing layers;

[0011] The content of each initial prediction image is modified based on the brain image to obtain each current prediction image.

[0012] Optionally, in the method according to the present invention, the brain image is regionally intercepted based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue, including:

[0013] Determining the tumor tissue body of each aneurysm tissue in the brain image respectively, and obtaining the tumor tissue contour corresponding to each tumor tissue body;

[0014] Determine in the brain image a first intersection point and a second intersection point where the contour of the aneurysm tissue corresponding to each aneurysm tissue is connected to the blood vessel, and determine along the extension direction of the blood vessel a first interception point and a second interception point that are spaced from the first intersection point and the second intersection point by a preset distance, respectively;

[0015] Determine the portion of the blood vessel between the first interception point and the second interception point as the blood vessel interception portion, and obtain each blood vessel interception portion corresponding to each aneurysm tissue;

[0016] The brain image is subjected to regional interception to obtain each tumor tissue region, wherein each tumor tissue region includes the tumor tissue body of each aneurysm tissue and each blood vessel interception portion corresponding to each aneurysm tissue.

[0017] Optionally, in the method according to the present invention, determining the corresponding predicted burst points based on the tissue morphology of each aneurysm tissue includes:

[0018] Connecting the first intersection points and the second intersection points respectively located in each tumor tissue region with line segments to form each tissue connection line, and determining the center points of each line segment located in each tissue connection line;

[0019] In each tumor tissue area, the edge contour points constituting the tumor tissue contour are determined, the distances between each edge contour point and the center point of the line segment are obtained respectively, and the edge contour point corresponding to the point with the largest distance is determined as the predicted burst point.

[0020] Optionally, in the method according to the present invention, generating a rupture prediction area in the image processing layer based on the predicted blasting point comprises:

[0021] Determine the connection length of the tissue connection line, and generate a rupture reference line with the same connection length on the tumor tissue contour with the predicted blasting point as the center;

[0022] Determining the contour length of the tumor tissue contour, and determining the length ratio between the contour length and the connection length;

[0023] Retrieving a preset rupture interval, and determining a rupture angle based on a ratio of the preset rupture interval to the length;

[0024] A rupture prediction region is generated in the image processing layer based on the rupture reference line and the rupture angle, wherein the rupture prediction region includes a tissue rupture region having an extension relationship with the tumor tissue contour.

[0025] Optionally, in the method according to the present invention, generating a rupture prediction area in the image processing layer based on the rupture reference line and the rupture angle comprises:

[0026] Acquiring the area of ​​the region constituting the tumor tissue contour, and determining the rupture extension length based on the area of ​​the region, the connection length and the rupture angle;

[0027] Determine a first endpoint and a second endpoint located at both ends of the rupture reference line, and generate endpoint connection lines respectively connected to the first endpoint and the second endpoint;

[0028] Determine a connecting vertical line perpendicular to the end point connecting line, and determine the extending direction of the connecting vertical line as the rupture direction;

[0029] A first rupture edge line and a second rupture edge line are generated with the first end point and the second end point as starting points, respectively, and are extended along the rupture direction and the angles between the first rupture edge line and the connecting vertical line are the rupture angles, wherein the lengths of the first rupture edge line and the second rupture edge line are both rupture extension lengths;

[0030] A rupture connection line connecting the first rupture edge line and the second rupture edge line is generated based on the rupture reference line to form a rupture prediction area consisting of the end point connection line, the first rupture edge line, the second rupture edge line and the rupture connection line.

[0031] Optionally, in the method according to the present invention, obtaining the area of ​​the region constituting the tumor tissue contour, and determining the rupture extension length based on the area of ​​the region, the connection length and the rupture angle comprises:

[0032] Normalizing the area of ​​the region based on the area normalization value to obtain an area coefficient, and normalizing the connection length based on the length normalization value to obtain a length coefficient;

[0033] Comparing the rupture angle with a preset angle interval to determine an angle coefficient, each preset angle interval having a preset angle coefficient;

[0034] The rupture extension length is calculated by the following formula,

[0035]

[0036] Among them, S2 is the rupture extension length, M is the area, gm is the area normalized value, k m is the area weight value, S1 is the connection length, g s is the length normalized value, k s is the length weight value, θ is the rupture angle, k α is the angle weight value, and γ is the training adjustment parameter.

[0037] Optionally, in the method according to the present invention, the method further comprises:

[0038] Displaying the rupture extension length, and if it is determined that the medical end adjusts the rupture extension length, extracting the training adjustment parameter γ;

[0039] receiving adjustment data inputted by the medical end for the rupture extension length, and if it is determined that the adjustment data is to extend the rupture extension length, increasing the training adjustment parameter γ based on the adjustment data;

[0040] If it is determined that the adjustment data is to shorten the rupture extension length, the training adjustment parameter γ is reduced based on the adjustment data, and the training adjustment parameter γ is trained by the following formula:

[0041]

[0042] Among them, β is the training adjustment parameter after training, S3 is the adjusted length corresponding to the adjusted data, and ρ + is the positive training coefficient, ρ - is the negative training coefficient.

[0043] Optionally, in the method according to the present invention, generating a fracture connection line respectively connecting the first fracture edge line and the second fracture edge line based on the fracture reference line comprises:

[0044] Determining an edge angle between the first rupture edge line or the second rupture edge line and the endpoint connection line, and determining a proportionality coefficient based on the edge angle;

[0045] Amplifying the rupture reference line based on a proportionality coefficient to obtain a rupture connection line;

[0046] The rupture connection line is connected to the first rupture edge line and the second rupture edge line respectively.

[0047] Optionally, in the method according to the present invention, each tumor tissue region and each corresponding rupture prediction region are image combined to obtain each initial prediction image with a rectangular outline located in each of the image processing layers, including:

[0048] Generate a reference rectangle that has a circumscribed relationship with the rupture prediction area and the tumor tissue area, and determine the rectangular center point of the reference rectangle;

[0049] Taking the center point of the rectangle as a reference, the reference rectangle is magnified by a preset multiple to obtain the rectangular outline;

[0050] The rupture prediction area and the tumor tissue area located in the rectangular outline are image combined to obtain initial prediction images with the rectangular outline as the image outline.

[0051] Optionally, in the method according to the present invention, performing content correction on each initial prediction image based on the brain image to obtain each current prediction image comprises:

[0052] Determine each predicted image center point located in each initial predicted image, and determine each brain image center point located in the brain image and corresponding to each predicted image center point;

[0053] Taking each brain image center point as the center, determining in the brain image corresponding image parts having the rectangular outline and corresponding to each initial prediction image, and extracting each corresponding image part;

[0054] Determine other regions except the rupture prediction region in each initial prediction image, and configure transparent pixel values ​​in the other regions to form each upper layer image;

[0055] Each corresponding image part is determined as a base layer, and an upper layer image corresponding to each corresponding image part is determined as a cover layer. The cover layer is covered on the corresponding base layer to obtain each current predicted image.

[0056] According to another aspect of the present invention, there is provided a digital-based intracranial aneurysm positioning device, comprising:

[0057] An acquisition module is configured to acquire a brain image of a patient, identify the brain image, and determine each aneurysmal tissue present in the brain image;

[0058] A capture module is configured to perform regional capture of the brain image based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue;

[0059] An establishing module is configured to establish image processing layers corresponding to the tumor tissue regions, and fill the tumor tissue regions into the corresponding image processing layers;

[0060] A prediction module, configured to determine corresponding predicted bursting points based on the tissue morphology of each aneurysm tissue, and generate a rupture prediction area in the image processing layer based on the predicted bursting points;

[0061] A combination module is configured to respectively combine each tumor tissue region with each corresponding rupture prediction region to obtain each initial prediction image with a rectangular outline located in each of the image processing layers;

[0062] The correction module is configured to perform content correction on each initial prediction image based on the brain image to obtain each current prediction image.

[0063] According to the solution of the present invention, the server can first obtain the patient's brain image, and identify the brain image, and identify each aneurysm tissue in the brain image; then, based on each aneurysm tissue, the brain image is intercepted, and each tumor tissue area including each aneurysm tissue is obtained after the image interception; then, each image processing layer corresponding to each tumor tissue area is established, and each tumor tissue area is filled into the corresponding image processing layer; then, based on the tissue morphology of each aneurysm tissue, the corresponding predicted blasting points are determined, and the rupture prediction area is generated in the image processing layer based on the predicted blasting point; at this time, the server will respectively combine each tumor tissue area with the corresponding rupture prediction area to obtain each initial prediction image located in each image processing layer; then, based on the brain image, the blank area located in each initial prediction image is corrected to obtain each current prediction image. The present invention can make a rough prediction of the rupture range and situation of the patient's intracranial aneurysm, and provide a certain reference for doctors. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A flow chart of a digital-based intracranial aneurysm positioning method according to an embodiment of the present invention is shown;

[0065] Figure 2 A schematic diagram showing the aneurysm tissue that has not been ruptured in this embodiment;

[0066] Figure 3 A schematic diagram of aneurysm tissue after rupture in this embodiment is shown;

[0067] Figure 4 A structural block diagram of a digital-based intracranial aneurysm positioning device according to another embodiment of the present invention is shown. DETAILED DESCRIPTION

[0068] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0069] Intracranial aneurysms are mostly abnormal bulges on the walls of intracranial arteries. They are the leading cause of subarachnoid hemorrhage and the third most common cause of cerebrovascular accidents, after cerebral thrombosis and hypertensive cerebral hemorrhage. They can occur at any age, but are more common in middle-aged and elderly women between 40 and 60 years old. For the treatment of intracranial aneurysms, a clear diagnosis must first be made through cerebral angiography, followed by early surgery and interventional embolization, because once an intracranial aneurysm ruptures, the mortality rate is extremely high.

[0070] The inventors found in their research that the existing method for locating intracranial aneurysms is usually performed by doctors by checking the patient's cerebral angiography, which lacks digital equipment to assist and has certain limitations.

[0071] In order to solve the problems existing in the above-mentioned prior art, the inventor proposes the solution of the present invention. An embodiment of the present invention provides a digital-based intracranial aneurysm positioning method, which can be executed in a computing device.

[0072] Figure 1 A flowchart of a digital-based intracranial aneurysm positioning method according to an embodiment of the present invention is shown. The method is suitable for execution in a computing device.

[0073] like Figure 1 As shown, the digital-based intracranial aneurysm positioning method in this embodiment starts with step S101. In step S101, the following contents are included:

[0074] A brain image of the patient is acquired, and the brain image is identified to determine each aneurysmal tissue present in the brain image.

[0075] For example, in this embodiment, the patient can obtain a brain image through CT scanning or the like, and the obtained brain image will be uploaded to the server. After receiving the brain image, the server will identify each blood vessel in the brain image, identify whether there is an aneurysm on the blood vessel, and further determine each aneurysm tissue present in the brain image; thereby achieving the effect of online identification, and further providing a certain auxiliary medical effect, so as to help medical staff improve the corresponding diagnostic accuracy and improve the corresponding work efficiency.

[0076] In step S102, the following contents are included:

[0077] The brain image is intercepted based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue.

[0078] For example, in this embodiment, after identifying each aneurysm tissue present in the brain image, the server will perform image capture of the area where these aneurysm tissues are located in the brain image, thereby obtaining each tumor tissue area including each aneurysm tissue; here, by performing image capture based on the aneurysm tissue, the brain image can be correspondingly noise removed, that is, the biological tissues other than the aneurysm tissues in the brain image can be removed on a large scale, and then in the subsequent prediction process, not only can a large amount of useless workload due to the presence of noise be avoided, but also the corresponding prediction accuracy can be improved.

[0079] Furthermore, the above-mentioned “intercepting the brain image based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue” further includes the following steps:

[0080] Determining the tumor tissue contours of each aneurysm tissue in the brain image;

[0081] Determine in the brain image a first intersection point and a second intersection point where the aneurysm tissue contour of each aneurysm tissue is connected to the blood vessel, and determine along the extension direction of the blood vessel a first interception point and a second interception point that are spaced from the first intersection point and the second intersection point by a preset distance respectively;

[0082] Determine the portion of the blood vessel between the first interception point and the second interception point as the blood vessel interception portion, and obtain each blood vessel interception portion corresponding to each aneurysm tissue;

[0083] The brain image is intercepted to obtain each tumor tissue region, wherein each tumor tissue region includes the tumor tissue contour of each aneurysm tissue and each blood vessel interception portion corresponding to each aneurysm tissue.

[0084] For example, in this embodiment, the contour recognition model can be obtained by performing corresponding model training on a pre-built image recognition model. Before performing model training, a corresponding contour training sample set can be obtained in advance. The contour training sample set can include a plurality of external contours corresponding to different aneurysm tissues, that is, contour training images of the contour of the aneurysm tissue. After obtaining the corresponding contour training sample set, the pre-built image recognition model can be trained by the contour training sample set to obtain the contour recognition model.

[0085] After identifying each aneurysm tissue present in the patient's brain image, the server will first determine the tumor tissue contour of each aneurysm tissue in the brain image through the contour recognition model. Then the server will determine the two intersection points where the tumor tissue contour of each aneurysm tissue is connected to the blood vessel in the brain image, namely the first intersection point and the second intersection point, and locate the position at a preset distance from the first intersection point and the second intersection point along the extension direction of the blood vessel, and then set the two positions as the first interception point and the second interception point respectively. The preset distance can be freely set according to the actual situation, such as 2mm, and the first interception point and the second interception point do not exist in the aneurysm tissue.

[0086] After determining the positions of the first interception point and the second interception point, the server determines the portion of the blood vessel between the first interception point and the second interception point as the blood vessel interception portion. At this time, the server can obtain the blood vessel interception portions corresponding to the aneurysm tissues, and perform image interception of the brain image according to the blood vessel interception portions. After the image interception, each tumor tissue area is obtained, and each tumor tissue area includes the tumor tissue contour of each aneurysm tissue and the blood vessel interception portions corresponding to each aneurysm tissue.

[0087] This embodiment can capture the image of each area where aneurysm tissue exists in the brain image, and pre-set the first interception point and the second interception point on both sides of the first intersection point and the second intersection point when capturing the image, thereby ensuring the integrity of the aneurysm tissue after image capture and reducing the workload of the subsequent server to a certain extent.

[0088] In step S103, the following contents are included:

[0089] Image processing layers corresponding to the tumor tissue regions are established, and the tumor tissue regions are filled into the corresponding image processing layers.

[0090] For example, in this embodiment, after obtaining each tumor tissue area, the server will respectively establish each image processing layer corresponding to the tumor tissue area. The image processing layer can be understood as a blank layer used to fill the tumor tissue area, and the number of layers of each blank layer established is the same as the number of areas of each tumor tissue area; after completing the establishment of each blank layer, the server will fill each tumor tissue area into the corresponding image processing layers.

[0091] In step S104, the following contents are included:

[0092] The corresponding predicted bursting points are determined based on the tissue morphology of each aneurysm tissue, and the rupture prediction area is generated in the image processing layer based on the predicted bursting points.

[0093] For example, in this embodiment, the server can identify the size and other tissue morphology of each aneurysm tissue, and determine the location where rupture may occur in the aneurysm tissue based on the identified tissue morphology, that is, the predicted rupture point, and then generate a rupture prediction area in the image processing layer based on the location of the predicted rupture point. Here, since this embodiment predicts the location of each predicted rupture point according to the tissue morphology of each aneurysm tissue, it has a certain diagnostic reference.

[0094] like Figure 2 as well as Figure 3 As shown, Figure 2 A schematic diagram showing unruptured aneurysm tissue. Figure 3 The corresponding diagram shows the ruptured aneurysm tissue.

[0095] Furthermore, the above-mentioned “determining the corresponding predicted burst points based on the tissue morphology of each aneurysm tissue” further includes the following steps:

[0096] Connecting the first intersection points and the second intersection points respectively located in each tumor tissue region with line segments to form each tissue connection line, and determining the center points of each line segment located in each tissue connection line;

[0097] In each tumor tissue area, the edge contour points constituting the tumor tissue contour are determined, the distances between each edge contour point and the center point of the line segment are obtained respectively, and the edge contour point corresponding to the point with the largest distance is determined as the predicted blasting point.

[0098] For example, in this embodiment, after determining the first intersection point and the second intersection point in each tumor tissue area, the server will connect the first intersection point and the second intersection point with a line segment, and the line segment formed by connecting the first intersection point and the second intersection point is the tissue connection line. At this time, the server will determine the center point of each tissue connection line, that is, the center point of each line segment. Then the server will determine each point in the tumor tissue contour in each tumor tissue area, that is, the edge contour point, and obtain the distance between each edge contour point and the center point of the line segment. According to the prior art, when the point distance is larger, it means that the position of the edge contour point is farther from the center point of the line segment, so the probability of aneurysm tissue rupture from the position of the edge contour point is higher; based on this feature, the server will determine the edge contour point corresponding to the largest point distance, and then determine the edge contour point as the predicted blasting point, thereby improving the corresponding prediction accuracy.

[0099] It should be noted that this embodiment can roughly predict the location where aneurysm tissue ruptures, and the location of the predicted rupture point is based on comparing the distances between each edge contour point and the center point of the line segment, and selecting the edge contour point corresponding to the largest point distance for prediction. The prediction process and prediction results are relatively objective and can provide medical staff with a certain reference value.

[0100] Furthermore, the above-mentioned “generating a rupture prediction area in the image processing layer based on the predicted blasting point” further includes the following steps:

[0101] Determine the connection length of the tissue connection line, and generate a rupture reference line with the same connection length on the tumor tissue contour with the predicted blasting point as the center;

[0102] Determining the contour length of the tumor tissue contour, and determining the length ratio between the contour length and the connection length;

[0103] Retrieving a preset rupture interval, and determining a rupture angle based on a ratio of the preset rupture interval to the length;

[0104] A rupture prediction region is generated in the image processing layer based on the rupture reference line and the rupture angle, wherein the rupture prediction region includes a tissue rupture region having an extension relationship with the tumor tissue contour.

[0105] For example, in this embodiment, after determining the predicted bursting point, the server will further determine the length of the tissue connection line, i.e., the connection length, and then generate a line segment with the same connection length on the contour of the tumor tissue, i.e., the rupture baseline, wherein the center point of the rupture baseline is the predicted bursting point; since this scheme is a scheme for roughly predicting the rupture range and rupture condition of an intracranial aneurysm, there may be a certain gap between the predicted result and the actual result, and in this embodiment, the center point of the generated rupture baseline is determined as the predicted bursting point, in order to correspondingly improve the accuracy of the prediction of the rupture range and rupture condition of an intracranial aneurysm, so as to provide medical staff with a certain reference.

[0106] Furthermore, the server will determine the length of the tumor tissue contour, that is, the contour length, and calculate the ratio between the contour length and the connection length, as well as the length ratio. The server will then retrieve the preset rupture interval, which can be manually set in advance based on real experience, and has different length ratios and rupture angles corresponding to the length ratios in the preset rupture interval. The rupture angle can be understood as the angle of the rupture range produced when the aneurysm tissue ruptures at the predicted burst point, and the server will then determine the rupture angle corresponding to the length ratio in the preset rupture interval.

[0107] After determining the rupture angle, the server will generate a rupture prediction area in the image processing layer according to the rupture baseline and the rupture angle. The area in the rupture prediction area that has an extension relationship with the aneurysm tissue contour is the tissue rupture area. This embodiment can manually pre-set the preset rupture interval based on real experience, so that the server can roughly predict the rupture range and rupture situation of the aneurysm tissue, which can save the calculation of the amount of data and quickly obtain the prediction results to provide a certain reference for medical staff.

[0108] Furthermore, the above-mentioned “generating a rupture prediction area in the image processing layer based on the rupture baseline and the rupture angle” further includes the following steps:

[0109] Acquiring a regional area of ​​a tissue region located in the tumor tissue contour, and determining a rupture extension length based on the regional area, the connection length, and the rupture angle;

[0110] Determine a first endpoint and a second endpoint located at both ends of the rupture reference line, and generate endpoint connection lines respectively connected to the first endpoint and the second endpoint;

[0111] Determine a connecting vertical line perpendicular to the end point connecting line, and determine the extending direction of the connecting vertical line as the rupture direction;

[0112] A first rupture edge line and a second rupture edge line are generated with the first end point and the second end point as starting points, respectively, and are extended along the rupture direction and the angles between the first rupture edge line and the connecting vertical line are the rupture angles, wherein the lengths of the first rupture edge line and the second rupture edge line are both rupture extension lengths;

[0113] A rupture connection line connecting the first rupture edge line and the second rupture edge line is generated based on the rupture reference line to form a rupture prediction area consisting of the end point connection line, the first rupture edge line, the second rupture edge line and the rupture connection line.

[0114] For example, in this embodiment, after determining the rupture reference line and the rupture angle, the server will first obtain the area of ​​the tissue area in the aneurysm tissue contour, and then determine the length of the rupture range generated when the aneurysm tissue ruptures at the predicted burst point based on the area, the connection length, and the rupture angle, that is, the rupture extension length. Then, the server will determine the two endpoints of the rupture reference line as the first endpoint and the second endpoint, and connect the first endpoint and the second endpoint with a straight line, that is, generate an endpoint connection line. At this time, the server will generate a line perpendicular to the endpoint connection line, that is, a connecting vertical line, and determine the extension direction of the connecting vertical line as the rupture direction.

[0115] After determining the rupture direction, the server will take the first endpoint and the second endpoint as the starting point and extend in the direction of the rupture direction, thereby generating the first rupture edge line and the second rupture edge line. The angle between the first rupture edge line and the second rupture edge line and the connecting vertical line is the determined rupture angle, and the lengths of the first rupture edge line and the second rupture edge line are both the determined rupture extension lengths. At this time, the server will generate a rupture connection line with the first rupture edge line and the second rupture edge line connected at both ends according to the rupture reference line, and form a rupture prediction area composed of the endpoint connection line, the first rupture edge line, the second rupture edge line and the rupture connection line.

[0116] The rupture prediction area generated in this embodiment is composed of endpoint connection lines, first rupture edge lines, second rupture edge lines and rupture connection lines, which can increase the convenience of medical staff in conducting corresponding inspections and save a certain amount of working time.

[0117] Furthermore, the above-mentioned “obtaining the area of ​​the tissue region located in the tumor tissue contour, and determining the rupture extension length based on the area of ​​the region, the connection length and the rupture angle” further includes the following steps:

[0118] Normalizing the area of ​​the region based on the area normalization value to obtain an area coefficient, and normalizing the connection length based on the length normalization value to obtain a length coefficient;

[0119] Comparing the rupture angle with a preset angle interval to determine an angle coefficient, each preset angle interval having a preset angle coefficient;

[0120] The rupture extension length is calculated by the following formula,

[0121]

[0122] Among them, S2 is the rupture extension length, M is the area, g m is the area normalized value, k m is the area weight value, S1 is the connection length, g s is the length normalized value, k s is the length weight value, θ is the rupture angle, k α is the angle weight value, and γ is the training adjustment parameter.

[0123] For example, in this embodiment, when it is necessary to determine the rupture extension length, it can be specifically performed in the following manner:

[0124] First, since the determination of the rupture extension length needs to be determined based on data in three different dimensions, namely, the area, the connection length, and the rupture angle, in order to achieve unified processing of the data in these three different dimensions, it is necessary to introduce corresponding normalization processing to process the data in the three different dimensions into data in the same dimension, that is, it is necessary to respectively retrieve the area normalization value and the length normalization value pre-stored in the server to respectively perform corresponding normalization processing on the area and the connection length, so as to obtain the corresponding area coefficient and length coefficient respectively;

[0125] Next, after the determination of the area coefficient and the length coefficient is completed, the rupture angle can be compared with the preset angle interval to determine the corresponding angle coefficient. It should be noted that the preset angle interval may include multiple ones, each preset angle interval includes different interval values, and each preset angle interval corresponds to a different angle coefficient. For example, when two preset angle intervals are included, the corresponding interval values ​​may be [10 degrees, 20 degrees], [30 degrees, 50 degrees], and the corresponding angle coefficients may be 0, 5, 0, 8. The rupture angle is compared with multiple preset angle intervals to determine the preset angle interval including the rupture angle, and further determine the angle coefficient corresponding to the preset angle interval.

[0126] Finally, after obtaining the area coefficient, length coefficient, and angle coefficient respectively, the corresponding rupture extension length can be calculated by the following formula:

[0127]

[0128] Among them, S2 is the rupture extension length, M is the area, g m is the area normalized value, k m is the area weight value, S1 is the connection length, g s is the length normalized value, k s is the length weight value, θ is the rupture angle, k α is the angle weight value, and γ is the training adjustment parameter.

[0129] It should be noted that, on the one hand, the above-mentioned area normalization value, area weight value, length normalization value, length weight value and angle weight value are all set by the medical staff themselves, and can be stored in the corresponding server after the setting is completed, and trigger the retrieval of the corresponding parameters when performing the above-mentioned calculations; on the other hand, after completing the corresponding prediction of the corresponding rupture extension length, if the medical staff believes that there is a certain gap between the predicted rupture extension length and the actual length, then it can be based on the above-mentioned training adjustment parameters. Make corresponding adjustments to achieve numerical adjustment of the formula, and then realize continuous adjustment of the prediction results, thereby improving the corresponding prediction accuracy.

[0130] For example, in this embodiment, the method may further include the following steps:

[0131] Displaying the rupture extension length, and if it is determined that the medical end adjusts the rupture extension length, extracting the training adjustment parameter γ;

[0132] receiving adjustment data inputted by the medical end for the rupture extension length, and if it is determined that the adjustment data is to extend the rupture extension length, increasing the training adjustment parameter γ based on the adjustment data;

[0133] If it is determined that the adjustment data is to shorten the rupture extension length, the training adjustment parameter γ is reduced based on the adjustment data, and the training adjustment parameter γ is trained by the following formula:

[0134]

[0135] Among them, β is the training adjustment parameter after training, S3 is the adjusted length corresponding to the adjusted data, and ρ + is the positive training coefficient, ρ - is the negative training coefficient.

[0136] For example, in this embodiment, the above-mentioned “displaying the rupture extension length” can be understood as displaying the predicted rupture extension length to the medical end for viewing. When the medical end views the corresponding rupture extension length, if it is considered that the predicted rupture extension length needs to be adjusted, the medical end will perform the corresponding adjustment process;

[0137] Furthermore, after receiving the corresponding adjustment processing operation, the server will extract the training adjustment parameters accordingly, and at the same time receive the adjustment data input by the medical terminal for the rupture extension length to perform corresponding adjustment processing;

[0138] On the one hand, if the server determines that the adjustment data is to extend the rupture extension length, the server will perform corresponding increase training on the training adjustment parameter based on the adjustment data, thereby completing the adjustment process of the training adjustment parameter;

[0139] On the other hand, if the server determines that the adjustment data is to shorten the rupture extension length, the server will reduce the training adjustment parameter based on the adjustment data, thereby completing the adjustment process of the training adjustment parameter;

[0140] Among them, the training of training adjustment parameters can be carried out in the following ways:

[0141]

[0142] For example, in this embodiment, after the training of the training adjustment parameters is completed through the above formula, the training adjustment parameters can be continuously updated and iterated, so that the prediction results become more accurate and the corresponding prediction effects are improved.

[0143] Furthermore, the above-mentioned “generating a rupture connection line respectively connecting the first rupture edge line and the second rupture edge line based on the rupture reference line” further includes the following steps:

[0144] Determining an edge angle between the first rupture edge line or the second rupture edge line and the endpoint connection line, and determining a proportionality coefficient based on the edge angle;

[0145] Amplifying the rupture reference line based on a proportionality coefficient to obtain a rupture connection line;

[0146] The rupture connection line is connected to the first rupture edge line and the second rupture edge line respectively.

[0147] For example, in this embodiment, after generating the first rupture edge line and the second rupture edge line, the server will determine the angle between the first rupture edge line or the second rupture edge line and the endpoint connection line, that is, the edge angle, and determine the proportional coefficient between the rupture baseline and the rupture connection line according to the edge angle. At this time, the server will enlarge the rupture baseline according to the proportional coefficient, and the line segment obtained after the enlargement is the rupture connection line, and then connect the two ends of the rupture connection line to the first rupture edge line and the second rupture edge line respectively. This embodiment can determine the proportional coefficient according to the edge angle, and then enlarge the rupture baseline according to the proportional coefficient to generate a rupture connection line, which has a certain objectivity and can provide a certain reference for medical staff.

[0148] In step S110, the following contents are included:

[0149] Each tumor tissue region and each corresponding rupture prediction region are image combined to obtain each initial prediction image located in each of the image processing layers.

[0150] For example, in this embodiment, after generating the rupture prediction region, the server will combine each tumor tissue region with its corresponding rupture prediction regions, that is, perform image combination. After the image combination, the image obtained by the server in the image processing layer is the initial prediction image. This embodiment can perform image combination of each tumor tissue region and the corresponding rupture prediction region, so that the initial prediction image after the image combination is more complete, which can facilitate the corresponding viewing of medical staff and improve work efficiency to a certain extent.

[0151] Furthermore, the above-mentioned “combining images of each tumor tissue region with the corresponding rupture prediction region to obtain each initial prediction image located in each of the image processing layers” further includes the following steps:

[0152] Generate a reference rectangle that has a circumscribed relationship with the rupture prediction area and the tumor tissue area, and determine the rectangular center point of the reference rectangle;

[0153] Taking the center point of the rectangle as a reference, the reference rectangle is magnified by a preset multiple to obtain a rectangular outline;

[0154] The rupture prediction area and the tumor tissue area located in the rectangular outline are image combined to obtain initial prediction images with the rectangular outline as the image outline.

[0155] For example, in this embodiment, after generating the rupture prediction region, the server will first generate a rectangle that has a circumscribed relationship with the rupture prediction region and the tumor tissue region, that is, a reference rectangle, and determine the rectangular center point of the reference rectangle. Then, based on the rectangular center point as a reference, the reference rectangle is enlarged by a preset multiple, and the size of the enlarged reference rectangle can fill the rupture prediction region and the tumor tissue region, thereby obtaining a rectangular outline. At this time, the server will combine the images of the rupture prediction region and the tumor tissue region located in the rectangular outline to obtain each initial prediction image with the rectangular outline as the image outline. This embodiment can make the image outline of the initial prediction image not too large or too small, and save a certain amount of workload on the server on the basis of being able to clearly and completely view the initial prediction image.

[0156] In step S112, the following contents are included:

[0157] Based on the brain image, the blank areas in each initial prediction image are respectively modified to obtain each current prediction image.

[0158] For example, in this embodiment, after obtaining the initial predicted images, the server will fill the blank areas in each initial predicted image according to the brain image, that is, perform content correction, and obtain each current predicted image after the content correction.

[0159] Furthermore, the above-mentioned “based on the brain image, respectively correcting the content of the blank areas in each initial prediction image to obtain each current prediction image” further includes the following steps:

[0160] Determine each predicted image center point located in each initial predicted image, and determine each brain image center point located in the brain image and corresponding to each predicted image center point;

[0161] Taking each brain image center point as the center, determining in the brain image each correction part having a rectangular outline and corresponding to each initial prediction image;

[0162] Each corrected part is compared with the corresponding initial predicted image, and each image area in each corrected part that has a corresponding relationship with each blank area is determined. Based on each image area, the content of each blank area is corrected to obtain each current predicted image.

[0163] For example, in this embodiment, after obtaining each initial predicted image, the server will determine the center point of each initial predicted image, that is, the center point of each predicted image, and determine the position of the center point of each predicted image in the brain image, that is, the center point of each brain image. At this time, the server will determine the part corresponding to the initial predicted image with a rectangular outline as the center in the brain image, that is, the correction part. Then, the server will compare the correction part with the initial predicted image, determine the image area in the correction part that has a corresponding relationship with the blank area, and fill these image areas into the blank area, that is, perform content correction, and obtain the current predicted image after the content correction.

[0164] This embodiment can combine the initial predicted image with other image areas in the brain image according to the relative position of the initial predicted image in the brain image, so as to obtain the current predicted image. The current predicted image is more complete than the initial predicted image, which can facilitate corresponding viewing by medical staff.

[0165] According to the solution of the present invention, the server can first obtain the patient's brain image, and identify the brain image, and identify each aneurysm tissue in the brain image; then, based on each aneurysm tissue, the brain image is intercepted, and each tumor tissue area including each aneurysm tissue is obtained after the image interception; then, each image processing layer corresponding to each tumor tissue area is established, and each tumor tissue area is filled into the corresponding image processing layer; then, based on the tissue morphology of each aneurysm tissue, the corresponding predicted blasting points are determined, and the rupture prediction area is generated in the image processing layer based on the predicted blasting point; at this time, the server will respectively combine each tumor tissue area with the corresponding rupture prediction area to obtain each initial prediction image located in each image processing layer; then, based on the brain image, the blank area located in each initial prediction image is corrected to obtain each current prediction image. The present invention can make a rough prediction of the rupture range and situation of the patient's intracranial aneurysm, and provide a certain reference for doctors.

[0166] Another embodiment of the present invention provides a digital intracranial aneurysm positioning device. Figure 4The corresponding device block diagram includes:

[0167] An acquisition module is configured to acquire a brain image of a patient, identify the brain image, and determine each aneurysmal tissue present in the brain image;

[0168] A capture module is configured to capture the brain image based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue;

[0169] An establishing module is configured to establish image processing layers corresponding to the tumor tissue regions, and fill the tumor tissue regions into the corresponding image processing layers;

[0170] A prediction module, configured to determine corresponding predicted bursting points based on the tissue morphology of each aneurysm tissue, and generate a rupture prediction area in the image processing layer based on the predicted bursting points;

[0171] A combination module is configured to respectively combine each tumor tissue region with each corresponding rupture prediction region to obtain each initial prediction image located in each of the image processing layers;

[0172] The correction module is configured to correct the content of the blank areas in each initial prediction image based on the brain image to obtain each current prediction image.

[0173] In the description provided herein, algorithms and displays are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the examples of the present invention. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to implement the content of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the preferred embodiment of the present invention.

[0174] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0175] Similarly, it should be understood that in order to streamline the present disclosure and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof.

[0176] Those skilled in the art will appreciate that the modules or units or components of the devices in the examples disclosed herein may be arranged in the devices described in the embodiment, or alternatively may be located in one or more devices different from the devices in the examples. The modules in the foregoing examples may be combined into one module or may be divided into multiple submodules.

[0177] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into multiple sub-modules or sub-units or sub-components. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0178] Furthermore, those skilled in the art will appreciate that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present invention and to form different embodiments.

[0179] In addition, some of the embodiments are described herein as methods or combinations of method elements that can be implemented by a processor of a computer system or by other devices that perform the functions. Therefore, a processor with necessary instructions for implementing the method or method elements forms a device for implementing the method or method elements. In addition, the elements described herein of the device embodiments are examples of devices for implementing the functions performed by the elements for the purpose of implementing the invention.

[0180] As used herein, unless otherwise specified, the use of ordinal numbers "first," "second," "third," etc. to describe common objects merely indicates that different instances of similar objects are involved, and is not intended to imply that the objects so described must have a given order in time, space, order, or in any other manner.

[0181] Although the present invention has been described according to a limited number of embodiments, it will be apparent to those skilled in the art, with the benefit of the above description, that other embodiments may be envisioned within the scope of the invention thus described. In addition, it should be noted that the language used in this specification is selected primarily for readability and teaching purposes, rather than for explaining or defining the subject matter of the present invention. Therefore, many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the appended claims. The disclosure of the present invention is illustrative and not restrictive with respect to the scope of the present invention, and the scope of the present invention is defined by the appended claims.

Claims

1. A digital intracranial aneurysm localization method, characterized in that: The following steps are involved: Acquiring a brain image of the patient, and identifying the brain image to determine each aneurysmal tissue present in the brain image; Performing regional interception on the brain image based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue; Establishing image processing layers corresponding to the tumor tissue regions, and filling the tumor tissue regions into the corresponding image processing layers; Determining corresponding predicted bursting points based on the tissue morphology of each aneurysm tissue, and generating a rupture prediction area in the image processing layer based on the predicted bursting points; Combining images of each tumor tissue region with the corresponding rupture prediction region to obtain initial prediction images with rectangular outlines located in each of the image processing layers; performing content correction on each initial prediction image based on the brain image to obtain each current prediction image; Generating a rupture prediction area in the image processing layer based on the predicted blasting point includes: Determine the connection length of the tissue connection line, and generate a rupture reference line with the same connection length on the tumor tissue contour with the predicted blasting point as the center; Determining the contour length of the tumor tissue contour, and determining the length ratio between the contour length and the connection length; Retrieving a preset rupture interval, and determining a rupture angle based on a ratio of the preset rupture interval to the length; Generate a rupture prediction region in the image processing layer based on the rupture baseline and the rupture angle, wherein the rupture prediction region includes a tissue rupture region having an extension relationship with the tumor tissue contour; Generating a rupture prediction area in the image processing layer based on the rupture reference line and the rupture angle includes: Acquiring the area of ​​the region constituting the tumor tissue contour, and determining the rupture extension length based on the area of ​​the region, the connection length and the rupture angle; Determine a first endpoint and a second endpoint located at both ends of the rupture reference line, and generate endpoint connection lines respectively connected to the first endpoint and the second endpoint; Determine a connecting vertical line perpendicular to the end point connecting line, and determine the extending direction of the connecting vertical line as the rupture direction; A first rupture edge line and a second rupture edge line are generated with the first end point and the second end point as starting points, respectively, and are extended along the rupture direction and the angles between the first rupture edge line and the connecting vertical line are the rupture angles, wherein the lengths of the first rupture edge line and the second rupture edge line are both rupture extension lengths; Based on the rupture reference line, a rupture connection line is generated to connect the first rupture edge line and the second rupture edge line respectively, so as to form a rupture prediction area consisting of the end point connection line, the first rupture edge line, the second rupture edge line and the rupture connection line; Acquiring the area of ​​the region constituting the tumor tissue contour, and determining the rupture extension length based on the area of ​​the region, the connection length, and the rupture angle, including: Normalizing the area of ​​the region based on the area normalization value to obtain an area coefficient, and normalizing the connection length based on the length normalization value to obtain a length coefficient; Comparing the rupture angle with a preset angle interval to determine an angle coefficient, each preset angle interval having a preset angle coefficient; The rupture extension length is calculated by the following formula, in, is the rupture extension length, is the area of ​​the region, is the area normalized value, is the area weight value, is the connection length, is the length normalized value, is the length weight value, is the rupture angle, is the angle weight value, Tune parameters for training.

2. The method for locating intracranial aneurysms based on digitization according to claim 1, characterized in that: Based on each aneurysm tissue, the brain image is regionally intercepted to obtain each tumor tissue region including each aneurysm tissue, including: Determining the tumor tissue body of each aneurysm tissue in the brain image respectively, and obtaining the tumor tissue contour corresponding to each tumor tissue body; Determine in the brain image a first intersection point and a second intersection point where the contour of the aneurysm tissue corresponding to each aneurysm tissue is connected to the blood vessel, and determine along the extension direction of the blood vessel a first interception point and a second interception point that are spaced from the first intersection point and the second intersection point by a preset distance, respectively; Determine the portion of the blood vessel between the first interception point and the second interception point as the blood vessel interception portion, and obtain each blood vessel interception portion corresponding to each aneurysm tissue; The brain image is subjected to regional interception to obtain each tumor tissue region, wherein each tumor tissue region includes the tumor tissue body of each aneurysm tissue and each blood vessel interception portion corresponding to each aneurysm tissue.

3. The method for locating intracranial aneurysms based on digitization according to claim 2, characterized in that: Determine the corresponding predicted burst points based on the tissue morphology of each aneurysm tissue, including: Connecting the first intersection points and the second intersection points respectively located in each tumor tissue region with line segments to form each tissue connection line, and determining the center points of each line segment located in each tissue connection line; In each tumor tissue area, the edge contour points constituting the tumor tissue contour are determined, the distances between each edge contour point and the center point of the line segment are obtained respectively, and the edge contour point corresponding to the point with the largest distance is determined as the predicted burst point.

4. The method for locating intracranial aneurysms based on digitization according to claim 1, characterized in that: The method further comprises: The rupture extension length is displayed, and if it is determined that the medical end adjusts the rupture extension length, the training adjustment parameter is adjusted. Perform extraction; receiving the adjustment data inputted by the medical end for the rupture extension length, and if it is determined that the adjustment data is to extend the rupture extension length, adjusting the training adjustment parameters based on the adjustment data; Conduct enlargement training; If the adjustment data is determined to be a shortening adjustment for the rupture extension length, the training adjustment parameter is adjusted based on the adjustment data. Perform reduction training and adjust the training parameters through the following formula train, in, Tune parameters for training after training, To adjust the length corresponding to the adjusted data, is the positive training coefficient, is the negative training coefficient.

5. The method for locating intracranial aneurysms based on digitization according to claim 1, characterized in that: Generating a fracture connection line respectively connecting the first fracture edge line and the second fracture edge line based on the fracture reference line includes: Determining an edge angle between the first rupture edge line or the second rupture edge line and the endpoint connection line, and determining a proportionality coefficient based on the edge angle; Amplifying the rupture reference line based on a proportionality coefficient to obtain a rupture connection line; The rupture connection line is connected to the first rupture edge line and the second rupture edge line respectively.

6. The method for locating intracranial aneurysms based on digitization according to claim 1, characterized in that: Combining each tumor tissue region with each corresponding rupture prediction region to obtain each initial prediction image with a rectangular outline located in each image processing layer, including: Generate a reference rectangle that has a circumscribed relationship with the rupture prediction area and the tumor tissue area, and determine the rectangular center point of the reference rectangle; Taking the center point of the rectangle as a reference, the reference rectangle is magnified by a preset multiple to obtain the rectangular outline; The rupture prediction area and the tumor tissue area located in the rectangular outline are image combined to obtain initial prediction images with the rectangular outline as the image outline.

7. The method for locating intracranial aneurysms based on digitization according to claim 6, characterized in that: Based on the brain image, each initial prediction image is modified in content to obtain each current prediction image, including: Determine each predicted image center point located in each initial predicted image, and determine each brain image center point located in the brain image and corresponding to each predicted image center point; Taking each brain image center point as the center, determining in the brain image corresponding image parts having the rectangular outline and corresponding to each initial prediction image, and extracting each corresponding image part; Determine other regions except the rupture prediction region in each initial prediction image, and configure transparent pixel values ​​in the other regions to form each upper layer image; Each corresponding image part is determined as a base layer, and an upper layer image corresponding to each corresponding image part is determined as a cover layer. The cover layer is covered on the corresponding base layer to obtain each current predicted image.

8. A digital intracranial aneurysm localization device, characterized in that: include: An acquisition module is configured to acquire a brain image of a patient, identify the brain image, and determine each aneurysmal tissue present in the brain image; A capture module is configured to perform regional capture of the brain image based on each aneurysm tissue to obtain each tumor tissue region including each aneurysm tissue; An establishing module is configured to establish image processing layers corresponding to the tumor tissue regions, and fill the tumor tissue regions into the corresponding image processing layers; A prediction module, configured to determine corresponding predicted bursting points based on the tissue morphology of each aneurysm tissue, and generate a rupture prediction area in the image processing layer based on the predicted bursting points; A combination module is configured to respectively combine each tumor tissue region with each corresponding rupture prediction region to obtain each initial prediction image with a rectangular outline located in each of the image processing layers; A correction module, configured to perform content correction on each initial prediction image based on the brain image to obtain each current prediction image; Generating a rupture prediction area in the image processing layer based on the predicted blasting point includes: Determine the connection length of the tissue connection line, and generate a rupture reference line with the same connection length on the tumor tissue contour with the predicted blasting point as the center; Determining the contour length of the tumor tissue contour, and determining the length ratio between the contour length and the connection length; Retrieving a preset rupture interval, and determining a rupture angle based on a ratio of the preset rupture interval to the length; Generate a rupture prediction region in the image processing layer based on the rupture baseline and the rupture angle, wherein the rupture prediction region includes a tissue rupture region having an extension relationship with the tumor tissue contour; Generating a rupture prediction area in the image processing layer based on the rupture reference line and the rupture angle includes: Acquiring the area of ​​the region constituting the tumor tissue contour, and determining the rupture extension length based on the area of ​​the region, the connection length and the rupture angle; Determine a first endpoint and a second endpoint located at both ends of the rupture reference line, and generate endpoint connection lines respectively connected to the first endpoint and the second endpoint; Determine a connecting vertical line perpendicular to the end point connecting line, and determine the extending direction of the connecting vertical line as the rupture direction; A first rupture edge line and a second rupture edge line are generated with the first end point and the second end point as starting points, respectively, and are extended along the rupture direction and the angles between the first rupture edge line and the connecting vertical line are the rupture angles, wherein the lengths of the first rupture edge line and the second rupture edge line are both rupture extension lengths; Based on the rupture reference line, a rupture connection line is generated to connect the first rupture edge line and the second rupture edge line respectively, so as to form a rupture prediction area consisting of the end point connection line, the first rupture edge line, the second rupture edge line and the rupture connection line; Acquiring the area of ​​the region constituting the tumor tissue contour, and determining the rupture extension length based on the area of ​​the region, the connection length, and the rupture angle, including: Normalizing the area of ​​the region based on the area normalization value to obtain an area coefficient, and normalizing the connection length based on the length normalization value to obtain a length coefficient; Comparing the rupture angle with a preset angle interval to determine an angle coefficient, each preset angle interval having a preset angle coefficient; The rupture extension length is calculated by the following formula, in, is the rupture extension length, is the area of ​​the region, is the area normalized value, is the area weight value, is the connection length, is the length normalized value, is the length weight value, is the rupture angle, is the angle weight value, Tune parameters for training.

Images