A method for preoperative detection of intracranial aneurysm embolization

Abstract

The application provides a preoperative detection method for intracranial aneurysm embolization, which comprises the following steps: acquiring a first intracranial image, analyzing the first intracranial image, determining a target lesion position, and determining a target influence range of embolization based on the target lesion position; marking and segmenting the target influence range in the first intracranial image to obtain a second intracranial image; acquiring position feature points of the second intracranial image, and constructing a three-dimensional simulation model based on the position feature points; analyzing the three-dimensional simulation model, and completing the preoperative detection of the intracranial aneurysm embolization based on the analysis result. By analyzing the first intracranial image, determining the target lesion position, determining the target influence range according to the target lesion position, and segmenting to determine the second intracranial image, the analysis range is reduced, the efficiency of preoperative detection is improved, the three-dimensional simulation model of the second intracranial image is constructed, the preoperative detection is facilitated, and the accuracy of preoperative detection is improved.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a preoperative detection method for intracranial aneurysm embolization. Background Technology

[0002] When a normal peripheral artery dilates into a sac-like shape, it is called an aneurysm. The aneurysm wall is thinner and less elastic than the normal vessel wall, making it prone to rupture and bleeding, which can lead to hemorrhagic shock. If the aneurysm ruptures in an intracranial artery, the patient often presents with severe headache, limb paralysis, or even loss of consciousness. In addition, the blood flow in the aneurysm cavity is in a vortex state for a long time, which can easily lead to secondary thrombosis. Therefore, peripheral aneurysms are extremely dangerous if left untreated.

[0003] With the continuous development of medical technology, aneurysm embolization has gradually become an important way to treat peripheral aneurysms. During aneurysm interventional embolization, microcatheters are often used to be superselectively inserted into the aneurysm cavity. Coils are delivered through the microcatheter to fill the aneurysm cavity, thereby achieving the purpose of sealing the aneurysm. However, when filling the aneurysm cavity, the condition of the aneurysm can only be understood through real-time imaging. If a decision is wrong in the face of an emergency, it will have serious consequences.

[0004] Therefore, the present invention provides a preoperative detection method for intracranial aneurysm embolization. Summary of the Invention

[0005] This invention provides a preoperative detection method for intracranial aneurysm embolization, which involves acquiring and analyzing a first intracranial image to determine the location of the target lesion, then determining the target area of ​​influence based on the location of the target lesion, and segmenting the image to determine a second intracranial image, thereby narrowing the analysis scope and improving the efficiency of preoperative detection. By constructing a three-dimensional simulation model of the second intracranial image, preoperative detection can be facilitated and its accuracy improved.

[0006] This invention provides a preoperative detection method for intracranial aneurysm embolization, comprising:

[0007] Step 1: Acquire the first intracranial image and analyze it to determine the location of the target lesion, and determine the target area of ​​influence of the embolization procedure based on the location of the target lesion.

[0008] Step 2: Mark and segment the target's area of ​​influence in the first intracranial image to obtain the second intracranial image;

[0009] Step 3: Obtain the location feature points of the second intracranial image, and construct a three-dimensional simulation model based on the location feature points of the second intracranial image;

[0010] Step 4: Analyze the three-dimensional simulation model and complete the preoperative detection for intracranial aneurysm embolization based on the analysis results.

[0011] Preferably, a preoperative detection method for intracranial aneurysm embolization includes, in step 1, acquiring a first intracranial image and analyzing the first intracranial image to determine the location of the target lesion, including:

[0012] Acquire a first intracranial image, and simultaneously divide the first intracranial image into equal intervals to obtain a sub-first intracranial image;

[0013] The pixel distribution features of each sub-first intracranial image are determined, and the corresponding sub-feature vector of each sub-first intracranial image is determined based on the pixel distribution features of each sub-first intracranial image.

[0014] Select the mode sub-feature vector set from the sub-feature vector set, and lock the remaining sub-feature vectors excluding the mode sub-feature vector set. Use the remaining sub-feature vector set as the target set. In the mode sub-feature vector set, each sub-feature vector is consistent.

[0015] The sub-first intracranial image corresponding to each sub-feature vector in the target set is determined, and each sub-first intracranial image is mapped into the first intracranial image. At the same time, the target lesion and the location of the target lesion corresponding to the target lesion are determined based on the mapping result.

[0016] Preferably, a preoperative detection method for intracranial aneurysm embolization, after determining the location of the target lesion based on the mapping results, further includes:

[0017] The target lesion is read to determine its first image features;

[0018] Extract target lesion samples from a pre-defined image library and obtain the second image features of the target lesion samples;

[0019] Obtain the target similarity between the first image feature and the second image feature, and compare the target similarity with a preset similarity to determine whether the target lesion is correct;

[0020] When the target similarity is equal to or greater than the preset similarity, the target lesion is determined to be correct;

[0021] Otherwise, the target lesion is determined to be incorrect, and the target lesion is re-identified.

[0022] Preferably, a preoperative detection method for intracranial aneurysm embolization, wherein step 1, determining the target area of ​​influence for embolization based on the location of the target lesion, is characterized by comprising:

[0023] The location of the target lesion is read in the first intracranial image, the shape of the lesion corresponding to the location of the target lesion is obtained, and the regional features corresponding to the location of the target lesion are determined based on the shape of the lesion.

[0024] Retrieve the target text of the embolization procedure from the preset surgical management database, and determine the execution characteristics of the embolization procedure based on the target text;

[0025] Based on regional and operational characteristics, the minimum safe distance between the normal area and the target lesion location is determined, and the target impact range of the embolization procedure is determined based on the minimum safe distance.

[0026] Preferably, in a preoperative detection method for intracranial aneurysm embolization, step 2 involves marking and segmenting the target area in a first intracranial image to obtain a second intracranial image, including:

[0027] The target's influence range is read, the boundary contour pixels of the target's influence range are determined, and the boundary contour pixels of the target's influence range are marked in the first intracranial image.

[0028] Based on the annotation results, the image cutting points are determined. At the same time, the target's influence range is segmented in the first intracranial image according to the image cutting points.

[0029] Based on the segmentation results, a second intracranial image was obtained.

[0030] Preferably, in a preoperative detection method for intracranial aneurysm embolization, step 3 involves acquiring location feature points from a second intracranial image and constructing a three-dimensional simulation model based on these location feature points, including:

[0031] Acquire a second intracranial image, determine the relative position of the intracranial aneurysm and vascular tissue based on the second intracranial image, and determine the intracranial structural features based on the relative position;

[0032] The execution requirements for aneurysm embolization are obtained, and the number of feature points to be selected is determined based on the execution requirements. The location feature points are then determined based on the intracranial structural features according to the number of feature points to be selected.

[0033] A preset planar image spatial transformation network is invoked to perform two-dimensional to three-dimensional spatial transformation on the second intracranial image based on location feature points, thereby obtaining a stereoscopic intracranial image. Based on the stereoscopic intracranial image, the intracranial skull, vascular tissue, and aneurysm tissue are identified respectively.

[0034] The intracranial skull, vascular tissue, and aneurysm tissue were scanned respectively to obtain the first point cloud data, the second point cloud data, and the third point cloud data corresponding to the intracranial skull, vascular tissue, and aneurysm tissue. The first point cloud data, the second point cloud data, and the third point cloud data were then sequentially input into a deep learning network for first modeling, second modeling, and third modeling to obtain the first three-dimensional model, the second three-dimensional model, and the third three-dimensional model.

[0035] The first, second, and third 3D models are integrated, and feature points are registered based on the location feature points to obtain a 3D simulation model.

[0036] The system receives model docking requests from a pre-defined server, obtains model parameters of the 3D simulation model based on the docking requests, and determines the target type of the 3D simulation model based on the model parameters.

[0037] The target device that interfaces with the 3D simulation model is determined based on the target type. The detection process of the target device on the 3D simulation model is configured based on the model parameters of the 3D simulation model. Based on the configuration results, the construction and interface of the 3D simulation model are completed.

[0038] Preferably, a preoperative detection method for intracranial aneurysm embolization includes constructing and connecting a three-dimensional simulation model based on configuration results, comprising:

[0039] Historical 3D simulation models are retrieved from a pre-set historical database, and a data retrieval process is constructed based on the retrieval results. A detection request is then sent to the target device based on the data retrieval process.

[0040] Once the target device responds, a call identifier is assigned to the historical 3D simulation model based on the data call process, and call guidance information is sent to the target device based on the call identifier.

[0041] Based on the call guidance information, the first monitoring is performed on the calling process of the target device on the historical 3D simulation model, and after the target device completes the calling of the historical 3D simulation model, the second monitoring is performed on the processing process of the target device on the historical 3D simulation model, and the target monitoring results are obtained.

[0042] Compare the target detection results with the standard results corresponding to historical intracranial images;

[0043] If the target detection result is consistent with the standard result, the detection process between the target device and the 3D simulation model is deemed to be qualified; otherwise, the detection process between the target device and the 3D simulation model is deemed to be unqualified, and the detection process is optimized.

[0044] Preferably, in a preoperative detection method for intracranial aneurysm embolization, step 4 involves analyzing a three-dimensional simulation model, including:

[0045] The constructed three-dimensional simulation model is obtained and read. The first morphological features of the target lesion in the three-dimensional simulation model are determined. At the same time, the second morphological features other than the block where the target lesion is located are determined.

[0046] The target morphological parameters of the intracranial aneurysm are determined based on the first morphological feature and the second morphological feature, and the target morphological parameters are used as the first data. At the same time, the target blood flow dynamic data of the intracranial aneurysm are determined based on the three-dimensional simulation model, and the target blood flow dynamic data are used as the second data.

[0047] Obtain the patient's identity information, extract the patient's historical medical data from a preset medical database based on the identity information, and determine the first weight of the first data and the second weight of the second data based on the historical medical data.

[0048] The first data, the first weight, the second data, and the second weight are input into a preset aneurysm risk assessment model for evaluation to obtain the risk assessment value of intracranial aneurysm. Based on the risk assessment value, the risk level of intracranial aneurysm embolization is determined, and the target attention location is determined based on the risk assessment results.

[0049] Based on the risk level, a target surgical plan is matched from a preset surgical plan library, and the surgical steps in the target surgical plan corresponding to the target attention position are optimized based on the target attention position. Based on the optimization results, the final surgical plan for intracranial aneurysm embolization is obtained.

[0050] Preferably, a preoperative detection method for intracranial aneurysm embolization, which determines the target attention location based on risk assessment results, includes:

[0051] Obtain the risk assessment results for intracranial aneurysms, and determine the feasible risk assessment values ​​corresponding to different locations of intracranial aneurysms when performing embolization based on the risk assessment results. Note that the location of intracranial aneurysms requiring surgery during embolization is not unique.

[0052] Each executable risk assessment value is compared with a preset risk threshold, and based on the comparison results, abnormal surgical locations with executable risk assessment values ​​greater than the preset risk thresholds are identified, and these abnormal surgical locations are determined as target attention locations.

[0053] Preferably, a preoperative detection method for intracranial aneurysm embolization further includes:

[0054] Based on the analysis of the three-dimensional simulation model, the spatial characteristics of the target lesion are determined;

[0055] Based on the spatial characteristics of the target lesion and combined with the theoretical diffusion coefficient of the target lesion, the lesion volume diffusion function is constructed.

[0056] The risk assessment coefficient for intracranial aneurysm embolization is determined based on the lesion volume diffusion function to assess the change in lesion volume of the target lesion during embolization and the maximum permissible operation time for intracranial aneurysm embolization.

[0057] The risk level of performing the current intracranial aneurysm embolization procedure is assessed based on the risk assessment coefficient, and an assessment report is generated and transmitted to the report receiving terminal based on the assessment results.

[0058] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0059] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0060] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0061] Figure 1 This is a flowchart of a preoperative detection method for intracranial aneurysm embolization according to an embodiment of the present invention;

[0062] Figure 2 This is a flowchart of step 1 in a preoperative detection method for intracranial aneurysm embolization according to an embodiment of the present invention.

[0063] Figure 3 This is a flowchart of step 2 in a preoperative detection method for intracranial aneurysm embolization according to an embodiment of the present invention. Detailed Implementation

[0064] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0065] Example 1:

[0066] This embodiment provides a preoperative detection method for intracranial aneurysm embolization, such as... Figure 1 As shown, it includes:

[0067] Step 1: Acquire the first intracranial image and analyze it to determine the location of the target lesion, and determine the target area of ​​influence of the embolization procedure based on the location of the target lesion.

[0068] Step 2: Mark and segment the target's area of ​​influence in the first intracranial image to obtain the second intracranial image;

[0069] Step 3: Obtain the location feature points of the second intracranial image, and construct a three-dimensional simulation model based on the location feature points of the second intracranial image;

[0070] Step 4: Analyze the three-dimensional simulation model and complete the preoperative detection for intracranial aneurysm embolization based on the analysis results.

[0071] In this embodiment, the first intracranial image can be an intracranial image of the target patient obtained based on technologies such as CT.

[0072] In this embodiment, the target lesion location can be the location of the intracranial aneurysm in the target patient.

[0073] In this embodiment, the target area of ​​influence can be the region or area that was mistakenly treated during surgery around the intracranial aneurysm during intracranial aneurysm embolization.

[0074] In this embodiment, the second intracranial influence can be achieved by marking the target influence range and the target lesion location in the first intracranial image, and then segmenting the image region corresponding to the target influence range from the first intracranial image based on the marking results. The target image range includes the location of the intracranial aneurysm.

[0075] In this embodiment, the location feature points can be pixels in the second intracranial image that can represent the intracranial structural features or the positional relationship between the lesion and normal cells, such as edge contour pixels, intracranial blood vessel junction pixels, etc.

[0076] In this embodiment, the three-dimensional simulation model can be a three-dimensional simulation model determined by performing 3D modeling on a computer based on the location feature points of the second intracranial image.

[0077] In this embodiment, the analysis of the three-dimensional simulation model can be performed by computer to analyze the structure of the second intracranial image, determine the vascular characteristics near the intracranial aneurysm (such as vascular distribution, vascular direction, and vascular thickness), and thus determine whether there will be interfering factors during the intracranial aneurysm embolization procedure. Interfering factors may include situations that interfere with the normal progress of the intracranial aneurysm embolization procedure, such as the obstruction of the target lesion location by the vascular distribution or vascular direction.

[0078] The beneficial effects of the above technical solution are as follows: by acquiring a first intracranial image and analyzing it, the location of the target lesion can be determined. Then, the target influence range can be determined by the location of the target lesion, and the second intracranial image can be segmented to determine the second intracranial image, thereby narrowing the analysis range and improving the efficiency of preoperative detection. By constructing a three-dimensional simulation model of the second intracranial image, it is possible to provide convenience for preoperative detection and improve the accuracy of preoperative detection.

[0079] Example 2:

[0080] Based on Example 1, this example provides a preoperative detection method for intracranial aneurysm embolization. Step 1 involves acquiring a first intracranial image and analyzing it to determine the location of the target lesion, including:

[0081] Acquire a first intracranial image, and simultaneously divide the first intracranial image into equal intervals to obtain a sub-first intracranial image;

[0082] The pixel distribution features of each sub-first intracranial image are determined, and the corresponding sub-feature vector of each sub-first intracranial image is determined based on the pixel distribution features of each sub-first intracranial image.

[0083] Select the mode sub-feature vector set from the sub-feature vector set, and lock the remaining sub-feature vectors excluding the mode sub-feature vector set. Use the remaining sub-feature vector set as the target set. In the mode sub-feature vector set, each sub-feature vector is consistent.

[0084] The sub-first intracranial image corresponding to each sub-feature vector in the target set is determined, and each sub-first intracranial image is mapped into the first intracranial image. At the same time, the target lesion and the location of the target lesion corresponding to the target lesion are determined based on the mapping result.

[0085] In this embodiment, the sub-first intracranial image can be each intracranial image block obtained by dividing the first intracranial image into equal intervals, and is a part of the first intracranial image.

[0086] In this embodiment, the pixel distribution features can be the number of rows and columns of pixels in each sub-first intracranial image, as well as the brightness of different pixels.

[0087] In this embodiment, the sub-feature vector can be determined based on the pixel distribution characteristics and is used to characterize the image features of the corresponding sub-first intracranial image.

[0088] In this embodiment, the mode sub-feature vector set can be the sub-feature vector with the same value and the largest proportion of the total sub-feature vector data among all the sub-feature vectors corresponding to the first intracranial image.

[0089] In this embodiment, the target set is the sub-feature vector corresponding to all sub-first intracranial images after removing the mode sub-feature vector set.

[0090] In this embodiment, the mode sub-feature vector set can be any region in all sub-first intracranial images that can characterize the patient's normal intracranial region.

[0091] In this embodiment, the target lesion can be the diseased site of the patient as represented in the first intracranial image.

[0092] The beneficial effects of the above technical solution are: by acquiring the first intracranial image, segmenting and analyzing the acquired first intracranial image, the normal and abnormal areas in the first intracranial image can be accurately locked, and the target lesion and its location can be accurately and effectively judged based on the locking result, thereby facilitating the determination of the target image range based on the target lesion and its location, and improving the practicality of preoperative detection.

[0093] Example 3:

[0094] Based on Example 2, this example provides a preoperative detection method for intracranial aneurysm embolization, which, after determining the location of the target lesion based on the mapping results, further includes:

[0095] The target lesion is read to determine its first image features;

[0096] Extract target lesion samples from a pre-defined image library and obtain the second image features of the target lesion samples;

[0097] Obtain the target similarity between the first image feature and the second image feature, and compare the target similarity with a preset similarity to determine whether the target lesion is correct;

[0098] When the target similarity is equal to or greater than the preset similarity, the target lesion is determined to be correct;

[0099] Otherwise, the target lesion is determined to be incorrect, and the target lesion is re-identified.

[0100] In this embodiment, the first image feature may be the image characteristics corresponding to the target lesion, specifically the area of ​​the target lesion or other features.

[0101] In this embodiment, the preset image library is pre-set and used to store different target lesion samples, wherein the target lesion samples are obtained by extraction.

[0102] In this embodiment, the second image feature can be a feature corresponding to the target lesion sample, specifically a feature such as the area corresponding to the lesion in the target lesion sample.

[0103] In this embodiment, the target similarity can be used to characterize the degree of similarity between the first image feature and the second image feature. The greater the target similarity, the more similar the first image feature and the second image feature are.

[0104] In this embodiment, the preset similarity is set in advance.

[0105] The beneficial effects of the above technical solution are: by analyzing the target lesion and comparing the target lesion with the target lesion samples stored in the preset image library, the accuracy of the target lesion determination is improved by comparing the target lesion samples with the obtained target lesion samples, thereby ensuring the reliability and accuracy of the preoperative detection of intracranial aneurysm embolization.

[0106] Example 4:

[0107] Based on Example 1, this example provides a preoperative detection method for intracranial aneurysm embolization, such as... Figure 2 As shown, in step 1, determining the target area of ​​influence of the embolization procedure based on the location of the target lesion is characterized by including:

[0108] Step 101: Read the location of the target lesion in the first intracranial image, obtain the shape of the lesion corresponding to the location of the target lesion, and at the same time, determine the regional features corresponding to the location of the target lesion based on the shape of the lesion.

[0109] Step 102: Retrieve the target text of the embolization procedure from the preset surgical management library, and determine the execution characteristics of the embolization procedure based on the target text;

[0110] Step 103: Based on regional and operational characteristics, determine the minimum safe distance between the normal area and the target lesion location, and determine the target impact range of the embolization procedure based on the minimum safe distance.

[0111] In this embodiment, the regional features may be the shape and area of ​​the target lesion corresponding to the target lesion location.

[0112] In this embodiment, the preset surgical management library is pre-set and used to store the surgical content text corresponding to aneurysm embolization.

[0113] In this embodiment, the target text may be the execution steps included in the aneurysm embolization procedure and the execution content corresponding to each execution step.

[0114] In this embodiment, the execution features may include the specific execution sequence and execution speed during aneurysm embolization surgery.

[0115] In this embodiment, the minimum safe distance can be a representation of the minimum distance that needs to be maintained between the lesion location and the normal area during the aneurysm embolization procedure.

[0116] The beneficial effects of the above technical solution are as follows: by analyzing the location of the target lesion in the first intracranial image, the regional characteristics of the target lesion can be accurately and effectively analyzed. Secondly, based on the analysis results, the target text of the matching embolization procedure can be determined, and the target text can be analyzed to achieve accurate and effective analysis of the execution features. Finally, based on the regional features and execution features, the target influence range of the embolization procedure can be accurately and reliably analyzed, thus ensuring the accuracy and reliability of preoperative detection for intracranial aneurysm embolization.

[0117] Example 5:

[0118] Based on Example 1, this example provides a preoperative detection method for intracranial aneurysm embolization, such as... Figure 3 As shown, in step 2, the target's area of ​​influence is marked and segmented in the first intracranial image to obtain the second intracranial image, including:

[0119] Step 201: Read the target's influence range, determine the boundary contour pixels of the target's influence range, and mark the boundary contour pixels of the target's influence range in the first intracranial image;

[0120] Step 202: Determine the image cutting points based on the annotation results. At the same time, segment the target's influence range in the first intracranial image according to the image cutting points.

[0121] Step 203: Based on the segmentation results, obtain the second intracranial image.

[0122] In this embodiment, the boundary contour pixels can be multiple pixels contained within the boundary of the target image range.

[0123] In this embodiment, the image cutting point can be the segmentation point corresponding to the lines representing the target image range and the normal region in the first intracranial image.

[0124] The beneficial effects of the above technical solution are: by analyzing the target image range and accurately confirming the boundary line between the target image range and the normal area in the first intracranial image based on the analysis results, the target image range in the first intracranial image can be accurately segmented according to the boundary line, thereby enabling accurate and effective acquisition of the second intracranial image, which provides convenience and guarantee for preoperative detection of intracranial aneurysm embolization.

[0125] Example 6:

[0126] Based on Example 1, this example provides a preoperative detection method for intracranial aneurysm embolization. In step 3, location feature points of a second intracranial image are acquired, and a three-dimensional simulation model is constructed based on the location feature points of the second intracranial image, including:

[0127] Acquire a second intracranial image, determine the relative position of the intracranial aneurysm and vascular tissue based on the second intracranial image, and determine the intracranial structural features based on the relative position;

[0128] The execution requirements for aneurysm embolization are obtained, and the number of feature points to be selected is determined based on the execution requirements. The location feature points are then determined based on the intracranial structural features according to the number of feature points to be selected.

[0129] A preset planar image spatial transformation network is invoked to perform two-dimensional to three-dimensional spatial transformation on the second intracranial image based on location feature points, thereby obtaining a stereoscopic intracranial image. Based on the stereoscopic intracranial image, the intracranial skull, vascular tissue, and aneurysm tissue are identified respectively.

[0130] The intracranial skull, vascular tissue, and aneurysm tissue were scanned respectively to obtain the first point cloud data, the second point cloud data, and the third point cloud data corresponding to the intracranial skull, vascular tissue, and aneurysm tissue. The first point cloud data, the second point cloud data, and the third point cloud data were then sequentially input into a deep learning network for first modeling, second modeling, and third modeling to obtain the first three-dimensional model, the second three-dimensional model, and the third three-dimensional model.

[0131] The first, second, and third 3D models are integrated, and feature points are registered based on the location feature points to obtain a 3D simulation model.

[0132] The system receives model docking requests from a pre-defined server, obtains model parameters of the 3D simulation model based on the docking requests, and determines the target type of the 3D simulation model based on the model parameters.

[0133] The target device that interfaces with the 3D simulation model is determined based on the target type. The detection process of the target device on the 3D simulation model is configured based on the model parameters of the 3D simulation model. Based on the configuration results, the construction and interface of the 3D simulation model are completed.

[0134] In this embodiment, the relative position can be a representation of the location of the aneurysm in the blood vessel, specifically which segment of the blood vessel it is located in.

[0135] In this embodiment, the structural features can be those that characterize the intracranial aneurysm and its structural relationship with intracranial blood vessels and other tissues.

[0136] In this embodiment, the execution requirements may be the objectives to be achieved during aneurysm embolization and the surgical steps involved in performing the aneurysm embolization procedure.

[0137] In this embodiment, the feature point to be selected can be used as a reference or as an observation point during aneurysm embolization.

[0138] In this embodiment, the number of feature points to be selected can be the number of points that need to be referenced when performing aneurysm embolization.

[0139] In this embodiment, the preset planar image space transformation network is pre-set and is used to convert the planar image into the corresponding three-dimensional image.

[0140] In this embodiment, the stereoscopic intracranial image can be a three-dimensional stereoscopic image obtained by spatially transforming the second intracranial image through a preset planar image spatial transformation network.

[0141] In this embodiment, the first point cloud data, the second point cloud data, and the third point cloud data are respectively the point sets of the outer surfaces corresponding to the intracranial skull, vascular tissue, and aneurysm tissue.

[0142] In this embodiment, the first modeling, the second modeling, and the third modeling can be achieved by using a deep learning network to process the first point cloud data, the second point cloud data, and the third point cloud data corresponding to the intracranial skull, vascular tissue, and aneurysm tissue, respectively, so as to realize the corresponding modeling of the three and obtain the corresponding three-dimensional models.

[0143] In this embodiment, the first three-dimensional model, the second three-dimensional model, and the third three-dimensional model can be three-dimensional models corresponding to the intracranial skull, vascular tissue, and aneurysm tissue, respectively.

[0144] In this embodiment, the integration of the first three-dimensional model, the second three-dimensional model, and the third three-dimensional model can be achieved by organically combining the three models according to their positions within the skull, thereby realizing the construction of a three-dimensional simulation model.

[0145] In this embodiment, the preset server is pre-set to interface the constructed 3D simulation model with the corresponding analysis or processing equipment, thereby facilitating the analysis of the constructed 3D simulation model and realizing the corresponding preoperative detection.

[0146] In this embodiment, the model parameters can be the specifications of the three-dimensional simulation model and the type of the three-dimensional simulation model.

[0147] In this embodiment, the target type can be a category representing the entity corresponding to the three-dimensional simulation model, specifically a simulation model corresponding to organs such as the intracranial cavity or thoracic cavity.

[0148] In this embodiment, the target device can be a medical device that analyzes a three-dimensional simulation model.

[0149] The beneficial effects of the above technical solution are as follows: By analyzing the second intracranial image and selecting corresponding feature points in the second intracranial image based on the analysis results, the second intracranial image is converted into a corresponding three-dimensional intracranial image based on the feature points. By analyzing the three-dimensional intracranial image, the point cloud data corresponding to the intracranial skull, blood vessel tissue, and aneurysm tissue can be accurately and effectively acquired, which facilitates the construction of a three-dimensional simulation model. Finally, the acquired point cloud data is processed by a deep learning network to accurately and reliably construct the three-dimensional simulation model corresponding to the second intracranial image. The constructed three-dimensional simulation model is then connected to the corresponding target device, thereby enabling accurate and reliable analysis of the three-dimensional simulation model and ensuring the accuracy and reliability of preoperative detection for intracranial aneurysm embolization.

[0150] Example 7:

[0151] Based on Example 6, this example provides a preoperative detection method for intracranial aneurysm embolization, which completes the construction and docking of a three-dimensional simulation model based on the configuration results, including:

[0152] Historical 3D simulation models are retrieved from a pre-set historical database, and a data retrieval process is constructed based on the retrieval results. A detection request is then sent to the target device based on the data retrieval process.

[0153] Once the target device responds, a call identifier is assigned to the historical 3D simulation model based on the data call process, and call guidance information is sent to the target device based on the call identifier.

[0154] Based on the call guidance information, the first monitoring is performed on the calling process of the target device on the historical 3D simulation model, and after the target device completes the calling of the historical 3D simulation model, the second monitoring is performed on the processing process of the target device on the historical 3D simulation model, and the target monitoring results are obtained.

[0155] Compare the target detection results with the standard results corresponding to historical intracranial images;

[0156] If the target detection result is consistent with the standard result, the detection process between the target device and the 3D simulation model is deemed to be qualified; otherwise, the detection process between the target device and the 3D simulation model is deemed to be unqualified, and the detection process is optimized.

[0157] In this embodiment, the preset historical database is pre-set and used to store different types of intracranial aneurysm models.

[0158] In this embodiment, the historical 3D simulation model can be an intracranial aneurysm model under different conditions, stored in a preset historical database.

[0159] In this embodiment, the data retrieval process can be a docking process used to characterize the target device and the constructed three-dimensional simulation model.

[0160] In this embodiment, the calling identifier can be the target device corresponding to different historical 3D simulation models, that is, the identity information of different historical 3D simulation models.

[0161] In this embodiment, the invocation guidance information may be a prompt to the target device to invoke the corresponding historical 3D simulation model, so that the target device can invoke the corresponding historical 3D simulation model accordingly.

[0162] In this embodiment, the first monitoring may be the monitoring of the target device's use of the corresponding historical 3D simulation model, used to determine whether the target model is making accurate and reliable use of the historical 3D simulation model.

[0163] In this embodiment, the second monitoring can be used to monitor the target device's processing of the historical 3D simulation model after it has been invoked, in order to determine whether the target device can perform accurate and reliable detection of the historical 3D simulation model.

[0164] In this embodiment, the target monitoring result can be the monitoring result obtained after monitoring the application process and processing process of the target device on the historical three-dimensional simulation model.

[0165] In this embodiment, the standard result can be the actual risk level of the intracranial aneurysm corresponding to the historical three-dimensional simulation model, specifically the risk level of aneurysm rupture, etc.

[0166] The beneficial effects of the above technical solution are as follows: by retrieving the corresponding historical 3D simulation model from the preset historical database, and connecting the historical 3D simulation model with the corresponding target device, and by monitoring the connection process and the processing process of the target device on the corresponding historical 3D simulation model, the processing process of the target device on the 3D simulation model can be accurately and effectively verified, thereby ensuring the accuracy of the analysis of the constructed 3D simulation model and improving the reliability of preoperative detection.

[0167] Example 8:

[0168] Based on Example 1, this example provides a preoperative detection method for intracranial aneurysm embolization. In step 4, the three-dimensional simulation model is analyzed, including:

[0169] The constructed three-dimensional simulation model is obtained and read. The first morphological features of the target lesion in the three-dimensional simulation model are determined. At the same time, the second morphological features other than the block where the target lesion is located are determined.

[0170] The target morphological parameters of the intracranial aneurysm are determined based on the first morphological feature and the second morphological feature, and the target morphological parameters are used as the first data. At the same time, the target blood flow dynamic data of the intracranial aneurysm are determined based on the three-dimensional simulation model, and the target blood flow dynamic data are used as the second data.

[0171] Obtain the patient's identity information, extract the patient's historical medical data from a preset medical database based on the identity information, and determine the first weight of the first data and the second weight of the second data based on the historical medical data.

[0172] The first data, the first weight, the second data, and the second weight are input into a preset aneurysm risk assessment model for evaluation to obtain the risk assessment value of intracranial aneurysm. Based on the risk assessment value, the risk level of intracranial aneurysm embolization is determined, and the target attention location is determined based on the risk assessment results.

[0173] Based on the risk level, a target surgical plan is matched from a preset surgical plan library, and the surgical steps in the target surgical plan corresponding to the target attention position are optimized based on the target attention position. Based on the optimization results, the final surgical plan for intracranial aneurysm embolization is obtained.

[0174] In this embodiment, the first morphological feature may be a representation of the shape of the target lesion in the three-dimensional simulation model and its distribution in vascular tissue.

[0175] In this embodiment, the second morphological feature may be the shape corresponding to the normal area and its positional relationship with the target lesion.

[0176] In this embodiment, the target morphological parameter can be a parameter that characterizes specific aspects such as the shape of an intracranial aneurysm.

[0177] In this embodiment, the first data may be the morphological parameters of an intracranial aneurysm.

[0178] In this embodiment, the target blood flow dynamic data can be parameters such as blood flow velocity and blood flow direction in intracranial vascular tissue.

[0179] In this embodiment, the second data may be parameters such as the blood flow status of the intracranial aneurysm.

[0180] In this embodiment, the preset medical database is pre-set and used to store the historical medical information of different patients within a certain period of time.

[0181] In this embodiment, historical medical information can be the patient's past medical records during their hospital visits.

[0182] In this embodiment, the first weight and the second weight can be the proportions of the first data and the second data in the preoperative detection for aneurysm embolization, respectively.

[0183] In this embodiment, the pre-trained aneurysm risk assessment model is used to analyze the collected data, thereby ensuring that preoperative detection for aneurysm embolization is effectively performed.

[0184] In this embodiment, the risk assessment value is used to characterize the risk of intracranial aneurysm; the higher the risk assessment value, the higher the risk of intracranial aneurysm.

[0185] In this embodiment, the target attention location can be a location that needs special attention during aneurysm embolization.

[0186] In this embodiment, the preset surgical plan library is pre-set and used to store different surgical plans.

[0187] In this embodiment, the target surgical procedure can be a surgical procedure suitable for current aneurysm embolization procedures.

[0188] The beneficial effects of the above technical solution are as follows: By analyzing the constructed three-dimensional simulation model, the morphological parameters and blood flow dynamic data of intracranial aneurysms can be accurately and effectively obtained. Secondly, by using the patient's historical medical information, the weights of the target morphological parameters and target blood flow dynamic data of the aneurysm can be accurately and reliably determined, thereby enabling accurate and reliable analysis of intracranial aneurysm embolization based on the weights, ensuring the reliability of the understanding of the specific situation of the aneurysm. Finally, the surgical plan for aneurysm embolization can be optimized based on the analysis results, improving the practicality of preoperative detection for aneurysm embolization.

[0189] Example 9:

[0190] Based on Example 8, this example provides a preoperative detection method for intracranial aneurysm embolization, which determines the target attention location based on risk assessment results, including:

[0191] Obtain the risk assessment results for intracranial aneurysms, and determine the feasible risk assessment values ​​corresponding to different locations of intracranial aneurysms when performing embolization based on the risk assessment results. Note that the location of intracranial aneurysms requiring surgery during embolization is not unique.

[0192] Each executable risk assessment value is compared with a preset risk threshold, and based on the comparison results, abnormal surgical locations with executable risk assessment values ​​greater than the preset risk thresholds are identified, and these abnormal surgical locations are determined as target attention locations.

[0193] In this embodiment, the executable risk assessment value can be a risk factor characterizing different locations when performing embolization of intracranial aneurysms.

[0194] In this embodiment, the preset risk threshold is set in advance and is used to measure whether the risk at different locations exceeds the expected requirements.

[0195] In this embodiment, the abnormal surgical location can be a surgical location where the executable risk assessment value is greater than a preset risk threshold.

[0196] The beneficial effects of the above technical solution are: by analyzing the risk assessment results of intracranial aneurysms, the feasible risk assessment values ​​for different surgical locations can be accurately and reliably analyzed, thereby facilitating the accurate and effective identification of surgical locations that require special attention in intracranial aneurysms, and enabling the optimization of surgical plans based on the target attention locations, thus ensuring the practicality of preoperative testing.

[0197] Example 10:

[0198] Based on Example 1, it also includes:

[0199] Based on the analysis of the three-dimensional simulation model, the spatial characteristics of the target lesion are determined. Then, based on these spatial characteristics and the theoretical diffusion coefficient of the target lesion, a lesion volume diffusion function is constructed. This function is used to assess the change in lesion volume during embolization. Finally, based on the maximum permissible operative time for intracranial aneurysm embolization, a risk assessment coefficient for the procedure is calculated. The specific process includes:

[0200] The three-dimensional simulation model is analyzed to determine the spatial characteristics of the target lesion, and the edge points of the target lesion are determined based on the spatial characteristics. At the same time, the area of ​​the target lesion in each plane is obtained according to the edge points, and the plane with the largest area is selected as the reference plane.

[0201] Based on the biological characteristics of the target lesion, the theoretical diffusion coefficient of the target lesion is obtained, and based on the theoretical diffusion coefficient of the target lesion and the reference surface of the target lesion, a diffusion function is constructed to show how the lesion volume of the target lesion diffuses with the theoretical diffusion coefficient.

[0202]

[0203] Where V(ζ) represents the diffusion function of the target lesion volume as the theoretical diffusion coefficient diffuses; ζ represents the theoretical diffusion coefficient; a represents the minimum independent variable of the reference surface in the coordinate system; b represents the maximum independent variable of the reference surface in the coordinate system; f(x1) represents the upper surface function in the reference surface; f(x2) represents the lower surface function in the reference surface; x1 represents the independent variable of the upper surface function in the reference surface; x2 represents the independent variable of the lower surface function in the reference surface; dx represents the integral over the independent variable; H represents the perpendicular distance between the farthest vertex of the target lesion and the reference surface; i represents the current solid block excluding the target lesion; s i h represents the base area of ​​the i-th solid block; i This represents the height of the i-th solid block;

[0204] Based on the function of the lesion volume of the target lesion diffusing with the theoretical diffusion coefficient, the change in lesion volume of the target lesion during embolization is estimated.

[0205] The risk factor for intracranial aneurysm embolization is calculated based on the volume change of the lesion.

[0206]

[0207] Where μ represents the risk factor for intracranial aneurysm embolization; V vary λ represents the volume of the lesion; λ represents the theoretical speed of embolization; T represents the maximum permissible operation time for embolization; ξ represents the influence coefficient of patient physical characteristics, with a value range of (0.012, 0.013); P represents the historical probability of accidents during intracranial aneurysm embolization.

[0208] Obtain the parameter coefficients, compare the risk coefficients with the reference coefficients, and determine the risk level for intracranial aneurysm embolization based on the comparison results;

[0209] When the risk coefficient is less than the reference coefficient, the risk level of intracranial aneurysm embolization is determined to be low.

[0210] When the risk coefficient equals the reference coefficient, the risk level of intracranial aneurysm embolization is determined to be moderate.

[0211] Otherwise, the risk level of intracranial aneurysm embolization is determined to be high.

[0212] A risk level report is generated based on the risk level of intracranial aneurysm embolization and sent to the report receiving terminal.

[0213] In this embodiment, the historical accident probability can be determined based on big data to represent the number of failed intracranial aneurysm embolization cases out of the total number of cases.

[0214] In this embodiment, since the target lesion is mostly an irregular three-dimensional block, the volume of the target lesion is obtained by cutting and piecing. That is, the reference surface of the target lesion is used as the bottom surface, the vertical distance between the farthest vertex of the target lesion and the reference surface is used as the height, the regular three-dimensional block is calculated, and then the volume of multiple three-dimensional blocks other than the target lesion is subtracted to obtain the final volume of the target lesion.

[0215] In this embodiment, the reference coefficient can be obtained from multiple experiments and is used to measure the risk level of intracranial aneurysm.

[0216] In this embodiment, the physical characteristic influence coefficient can be the influence factors on the patient's age, gender, blood pressure, and other physical conditions during intracranial aneurysm surgery.

[0217] In this embodiment, the theoretical diffusion coefficient can be the coefficient of change of intracranial aneurysm over time, which is determined based on the rules of change of intracranial aneurysm. The theoretical diffusion coefficient changes with specific circumstances, but during the process of intracranial aneurysm embolization, the theoretical diffusion coefficient is a specific and constant value.

[0218] The beneficial effects of the above technical solution are: by accurately and effectively calculating the diffusion function of the target lesion corresponding to the intracranial aneurysm, and based on the calculation results, the risk coefficient of the intracranial aneurysm can be accurately and reliably calculated, which facilitates timely and effective understanding of the patient's current risk status of the intracranial aneurysm, thereby facilitating the selective formulation of appropriate surgery and improving the safety factor of arterial embolization.

[0219] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A preoperative detection method for intracranial aneurysm embolization, characterized in that, include: Step 1: Acquire the first intracranial image and analyze it to determine the location of the target lesion, and determine the target area of ​​influence of the embolization procedure based on the location of the target lesion. Step 2: Mark and segment the target's area of ​​influence in the first intracranial image to obtain the second intracranial image; Step 3: Obtain the location feature points of the second intracranial image, and construct a three-dimensional simulation model based on the location feature points of the second intracranial image; Step 4: Analyze the three-dimensional simulation model and complete the preoperative detection for intracranial aneurysm embolization based on the analysis results; This also includes: The three-dimensional simulation model is analyzed to determine the spatial characteristics of the target lesion, and the edge points of the target lesion are determined based on the spatial characteristics. At the same time, the area of ​​the target lesion in each plane is obtained according to the edge points, and the plane with the largest area is selected as the reference plane. Based on the biological characteristics of the target lesion, the theoretical diffusion coefficient of the target lesion is obtained, and based on the theoretical diffusion coefficient of the target lesion and the reference surface of the target lesion, a diffusion function is constructed to show how the lesion volume of the target lesion diffuses with the theoretical diffusion coefficient. ; in, The diffusion function represents the diffusion of the target lesion volume with respect to the theoretical diffusion coefficient; Indicates the theoretical diffusion coefficient; This represents the smallest independent variable of the reference plane in the coordinate system; This represents the maximum independent variable of the reference plane in the coordinate system; Represents the surface function on the reference plane; Represents the lower surface function in the reference plane; The independent variable represents the function of the upper surface in the reference plane; The independent variable represents the function of the lower surface in the reference plane; This indicates the integral over the independent variable; This represents the perpendicular distance between the farthest vertex from the reference plane in the target lesion and the reference plane. This represents the current 3D block excluding the target lesion; Let represent the base area of ​​the i-th solid block; This represents the height of the i-th solid block; Based on the function of the lesion volume of the target lesion diffusing with the theoretical diffusion coefficient, the change in lesion volume of the target lesion during embolization is estimated. The risk factor for intracranial aneurysm embolization is calculated based on the volume change of the lesion. ; in, This indicates the risk factor for intracranial aneurysm embolization. Indicates the change in volume of the lesion; Indicates the theoretical speed of embolization; indicates the maximum permissible operative time for performing embolization. This represents the influence coefficient of the patient's physical characteristics, and its value ranges from (0.012 to 0.013). Indicates the probability of historical accidents during intracranial aneurysm embolization; Obtain the parameter coefficients, compare the risk coefficients with the reference coefficients, and determine the risk level for intracranial aneurysm embolization based on the comparison results; When the risk coefficient is less than the reference coefficient, the risk level of intracranial aneurysm embolization is determined to be low. When the risk coefficient equals the reference coefficient, the risk level of intracranial aneurysm embolization is determined to be moderate. Otherwise, the risk level of intracranial aneurysm embolization is determined to be high; A risk level report is generated based on the risk level of intracranial aneurysm embolization and sent to the report receiving terminal.

Images