Search

CN-121982467-A - Method and system for identifying anatomical variation before breast endoscopic surgery

CN121982467ACN 121982467 ACN121982467 ACN 121982467ACN-121982467-A

Abstract

The invention discloses a method and a system for identifying anatomical variation before breast endoscopic surgery, and relates to the technical field of computer data processing. The method comprises the steps of S1, obtaining heterogeneous source data, carrying out space registration and data standardization fusion to generate a multidimensional fusion data set, S2, carrying out basic topology feature extraction based on the multidimensional fusion data set, calculating side weight information to construct a target topological graph, S3, carrying out matching and matching on the target topological graph and a preset standard topological database, calculating topology compliance parameters based on comparison differences, S4, identifying parameter abnormal areas according to space distribution features of the topology compliance parameters, and generating a space safety constraint model by combining local space medium attributes, and S5, outputting structure abnormality information and corresponding optimized path planning parameters based on the space safety constraint model. Through multichannel tensor quantization variation characteristics, a safety constraint is constructed by utilizing an anisotropic bounding box, and a potential field algorithm is combined to optimize a path, so that high-precision automatic obstacle avoidance is realized.

Inventors

  • SHANG FANGJIAN
  • WANG HAOBO
  • WANG YIFANG
  • LIU BO

Assignees

  • 河北医科大学第一医院

Dates

Publication Date
20260505
Application Date
20260206

Claims (10)

  1. 1. A method for identifying pre-operative anatomical variations of a breast endoscope, comprising: s1, acquiring at least two types of heterogeneous source data aiming at a target area, and carrying out space registration and data standardization fusion based on a preset global reference coordinate system to generate a multidimensional fusion data set; s2, extracting basic topological features based on the multidimensional fusion data set, and calculating side weight information to construct a target topological graph representing a structure communication relation in a target area; s3, matching and matching the target topological graph with a preset standard topological database, and calculating topological compliance parameters reflecting the deviation degree of the target area structure based on the comparison difference; s4, identifying a parameter abnormal region according to the spatial distribution characteristics of the topological compliance parameters, and generating a spatial safety constraint model with direction difference by combining local spatial medium attributes; S5, based on the space safety constraint model, outputting the structural abnormality information and the corresponding optimized path planning parameters.
  2. 2. The method for identifying pre-endoscopic anatomical variations according to claim 1, wherein said spatial registration in S1 specifically comprises: respectively acquiring a first static discrete data matrix of the same target object as a first mode and a second dynamic time sequence data stream as a second mode through a data transmission interface; Analyzing file header information of the first mode and the second mode, and extracting original resolution parameters and physical space mapping parameters; Executing weight distribution logic based on resolution value comparison, and respectively generating corresponding data source confidence weight values for the first modality and the second modality, wherein the higher the original resolution parameter is, the higher the source confidence weight value of the distributed data is; respectively traversing and calculating the data change rate numerical distribution of the first modality and the second modality in the three-dimensional logic space by utilizing a neighborhood numerical differential calculation rule, screening out discrete coordinate points with the change rate numerical value exceeding a preset threshold value, and constructing a first characteristic coordinate set and a second characteristic coordinate set; based on a minimum distance iterative optimization strategy, calculating Euclidean distances between the first feature coordinate set and the second feature coordinate set, and taking a space transformation matrix with the minimized Euclidean distances as a transformation operator for space dimension alignment.
  3. 3. The method for identifying pre-endoscopic anatomical variations according to claim 2, wherein the data normalization fusion in S1 specifically comprises: mapping the second mode to a coordinate system where the first mode is located by utilizing a transformation operator, and performing numerical interpolation calculation based on a space neighborhood on the mapped discrete data points to generate a standardized data volume; Constructing a multidimensional composite data tensor, mapping basic numerical value attributes of a standardized data volume to a first dimension, mapping data change rate numerical value distribution to a second dimension, and mapping corresponding source confidence weight to a third dimension to generate a multidimensional fusion data set.
  4. 4. The method for identifying pre-endoscopic anatomical variations according to claim 1, wherein the extracting of the basic topological feature in S2 specifically comprises: Based on the multi-dimensional fusion data set, extracting continuous high-dimensional data meeting preset connectivity conditions by utilizing numerical distribution characteristics mapped to the first dimension and the second dimension; and carrying out three-dimensional morphological refinement operation on the continuous high-dimensional data, and iteratively removing boundary voxel points on the premise of keeping the topological Euler number unchanged to generate a single element width central geometric skeleton set as a basic topological feature of the target region.
  5. 5. The method for identifying pre-endoscopic anatomical variations according to claim 4, wherein the constructing the target topology in S2 specifically comprises: Traversing each voxel point in the basic topological feature, calculating the number of connected components of the voxel points in the three-dimensional space neighborhood, marking the voxel points with the number of connected components larger than a preset threshold as bifurcation nodes, and marking the voxel points with the number of connected components equal to 1 as endpoint nodes; Tracking skeleton paths connecting the bifurcation nodes and the endpoint nodes, generating a topological edge set, and calculating space geometric attributes and medium density attributes corresponding to the topological edges based on the multidimensional fusion data set; and constructing a target topological graph by taking the bifurcation nodes and the endpoint nodes as graph nodes, taking the topological edge set as graph connecting edges and taking the space geometric attribute and the medium density attribute as edge weight information.
  6. 6. The method for identifying pre-operative anatomical variation of a breast endoscope according to claim 1, wherein the matching comparison in S3 specifically comprises: calling a preset standard topology database, wherein the standard topology database comprises a reference topology model generated based on a large number of historical sample statistics; performing sub-graph isomorphism matching operation on the target topological graph and the reference topological model by taking the edge weight information as a similarity measurement basis, and establishing a node mapping relation; based on the node mapping relationship, identifying redundant nodes which cannot be mapped in the target topological graph, and marking the redundant nodes as topological variation components.
  7. 7. The method for identifying pre-endoscopic anatomical variations according to claim 6, wherein the calculating topological compliance parameters in S3 specifically comprises: Calculating a space displacement vector of a corresponding reference node in a reference topological model based on the map node with the established mapping relation in the target topological graph; extracting a local gradient modulus value corresponding to the position of a graph node from a second dimension of the multidimensional fusion data set, and taking the local gradient modulus value as the physical rigidity weight of the graph node; Calculating the product of the modulus of the space displacement vector and the physical rigidity weight to generate a geometric strain energy component; And distributing a preset structural penalty weight for the topological variation component, and carrying out weighted fusion on the structural penalty weight and the geometric strain energy component to obtain the topological compliance parameter.
  8. 8. The method for identifying pre-endoscopic anatomical variations according to claim 1, wherein S4 specifically comprises: mapping the topological compliance parameters to a three-dimensional logic space corresponding to the multidimensional fusion data set, and constructing a potential energy density distribution field; traversing potential energy density distribution fields, and defining a space region with the numerical value exceeding a preset deviation threshold value as a parameter abnormal region; Based on the parameter anomaly region, extracting a corresponding local gradient main direction from a first dimension and a second dimension of the multi-dimensional fusion data set to serve as a direction difference guide vector; Taking the parameter abnormal region as a center, and carrying out space asymmetric expansion along the direction of the direction difference guide vector to generate an anisotropic bounding box set; and calculating the mutual exclusion influence range of each anisotropic bounding box in the three-dimensional logic space by using a space set union operation and a collision detection algorithm, and constructing a space safety constraint model.
  9. 9. The method for identifying pre-endoscopic anatomical variations according to claim 1, wherein S5 specifically comprises: taking the space safety constraint model as an impassable constraint, and constructing a global energy potential field taking the accumulated integral of the topological compliance parameter as a cost function in a three-dimensional logic space; Searching a discrete point sequence with the minimum cost value in the global energy potential field by using a heuristic search algorithm based on a preset initial coordinate and a target coordinate to serve as an original planning path; Performing curve fitting processing on the original planning path, extracting key control point coordinates and corresponding direction difference guide vectors, and generating optimized path planning parameters; Extracting topological characteristics and corresponding topological compliance parameters of the parameter abnormal region, and packaging the topological characteristics and the corresponding topological compliance parameters into structure abnormal information; and establishing an association relation between the structural abnormality information and the optimized path planning parameters, and executing asynchronous output in a structured data report form.
  10. 10. A breast endoscopic preoperative anatomical variation identification system for implementing a breast endoscopic preoperative anatomical variation identification method according to any one of claims 1-9, comprising: The multi-source data fusion module is used for acquiring at least two types of heterogeneous source data aiming at a target area, and carrying out space registration and data standardization fusion based on a preset global reference coordinate system to generate a multi-dimensional fusion data set; The topological structure modeling module is used for extracting basic topological features based on the multidimensional fusion data set, calculating side weight information and constructing a target topological graph representing the structure communication relation in the target area; Comparing the target topological graph with a preset standard topological database, and calculating topological compliance parameters reflecting the deviation degree of the target area structure based on comparison difference; The risk space constraint module is used for identifying a parameter abnormal region according to the space distribution characteristics of the topological compliance parameters and generating a space safety constraint model with direction difference by combining local space medium attributes; And the intelligent planning output module is used for outputting the structural abnormality information and the corresponding optimized path planning parameters based on the space safety constraint model.

Description

Method and system for identifying anatomical variation before breast endoscopic surgery Technical Field The invention relates to the technical field of computer data processing, in particular to a method and a system for identifying anatomical variation before breast endoscopic surgery. Background In many technical fields in which the interior of a closed or complex structure needs to be subjected to fine operation, the operation safety and success rate are highly dependent on accurate grasp of three-dimensional topological features in the interior of a target structure, for example, in a breast endoscopic operation, the accurate identification of an anatomical structure and potential anatomical variation in the target region has important significance in reducing operation risks and improving path planning safety. In practical operation, the internal tissue structure of the mammary gland region is complex, and the distribution of blood vessels and catheter systems is obviously different among different individuals. Currently, analysis of such complex structures is mostly dependent on associative analysis of multi-source heterogeneous probe data. However, it is difficult to digitally model complex three-dimensional connected structures by means of only single-modality discrete images, and it is difficult to automatically identify and quantify the risk of anatomical variation regions due to the lack of an effective spatial quantitative evaluation mechanism. When an auxiliary path planning model is constructed, if static anatomic data cannot be combined with dynamic space constraint logic, the generated planning path is difficult to avoid a high-risk variation area, so that the operation planning process still depends on more manual experience, and potential operation risks caused by structural irregularity are difficult to systematically predict and avoid. Therefore, how to realize effective fusion of multi-source data in a digital space and automatically identify topological structure variation based on structural features, and further construct a planning model with space constraint capability becomes a problem to be solved in the current data processing and auxiliary navigation fields. Disclosure of Invention Based on the above-mentioned drawbacks of the prior art, the present invention is directed to a method and a system for identifying anatomical variation before breast endoscopic surgery, so as to solve the above-mentioned technical problems. In order to achieve the purpose, the invention provides the following technical scheme that the method for identifying the anatomical variation before breast endoscopic surgery comprises the following steps: s1, acquiring at least two types of heterogeneous source data aiming at a target area, and carrying out space registration and data standardization fusion based on a preset global reference coordinate system to generate a multidimensional fusion data set; s2, extracting basic topological features based on the multidimensional fusion data set, and calculating side weight information to construct a target topological graph representing a structure communication relation in a target area; s3, matching and matching the target topological graph with a preset standard topological database, and calculating topological compliance parameters reflecting the deviation degree of the target area structure based on the comparison difference; s4, identifying a parameter abnormal region according to the spatial distribution characteristics of the topological compliance parameters, and generating a spatial safety constraint model with direction difference by combining local spatial medium attributes; S5, based on the space safety constraint model, outputting the structural abnormality information and the corresponding optimized path planning parameters. The present invention further provides that the spatial registration in S1 specifically includes: respectively acquiring a first static discrete data matrix of the same target object as a first mode and a second dynamic time sequence data stream as a second mode through a data transmission interface; Analyzing file header information of the first mode and the second mode, and extracting original resolution parameters and physical space mapping parameters; Executing weight distribution logic based on resolution value comparison, and respectively generating corresponding data source confidence weight values for the first modality and the second modality, wherein the higher the original resolution parameter is, the higher the source confidence weight value of the distributed data is; respectively traversing and calculating the data change rate numerical distribution of the first modality and the second modality in the three-dimensional logic space by utilizing a neighborhood numerical differential calculation rule, screening out discrete coordinate points with the change rate numerical value exceeding a preset threshold value, an