A method and related apparatus for generating surgical safety boundaries based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation.

Abstract

This application provides a method and related apparatus for generating surgical safety boundaries based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation. The method includes: S1. Obtaining a binary mask of key anatomical structures and their corresponding anatomical categories; S2. Performing time-varying settings on anisotropic parameters based on the anatomical category and the current surgical stage; S3. Performing anisotropic Euclidean distance transformation on the binary mask based on the anisotropic parameters; S4. Extracting multi-level safety isometric surfaces from the anisotropic distance field based on multi-level distance thresholds; S5. Encapsulating the multi-level safety isometric surfaces and complete metadata into a safety boundary data package; S6. Traceably exporting based on various standard 3D model file formats. This application also provides related apparatus corresponding to the method, including devices, electronic devices, computer-readable storage media, and computer program products.

Description

Technical Field

[0001] This application relates to the field of medical image processing, and in particular to a method and related apparatus for generating surgical safety boundaries based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation. Background Technology

[0002] In surgical planning and intraoperative navigation, the safety boundaries around key anatomical structures are a core visualization tool to help surgeons avoid iatrogenic injuries. Accurate generation of these safety boundaries is crucial for preoperative planning, intraoperative visualization, and postoperative assessment.

[0003] Existing methods for generating surgical safety boundaries have the following problems: First, they use isotropic distance transformation (i.e., the distance weights are the same in all directions), failing to consider the differences in clinical risk of different types of anatomical structures in different directions; second, the anisotropy parameters use fixed values ​​and are not dynamically adjusted according to the surgical stage; third, they only generate safety boundaries with a single distance threshold, failing to provide multi-level safety boundaries to support refined surgical planning; fourth, the safety isometric surface extraction algorithm is not robust enough on complex 3D structures; fifth, the output only contains geometric data and does not include complete metadata (anatomical category, parameters, volume statistics, etc.), making it difficult for downstream applications to understand and use; sixth, the output format is uniform, making it difficult to interface with different downstream surgical planning and navigation systems; seventh, there is a lack of linkage mechanism with preoperative surgical planning, resulting in an information gap between preoperative planning and safety boundary generation. Summary of the Invention

[0004] To address the above technical issues, this application provides a method and related apparatus for generating surgical safety boundaries based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation.

[0005] Firstly, a method for generating surgical safety boundaries based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation is provided, including the following steps: S1. Obtain a binary mask of the key anatomical structure and its corresponding anatomical category. The binary mask represents the occupancy of the key anatomical structure in three-dimensional space, and the anatomical category includes at least one of the following: blood vessel, nerve, and canal. S2. Based on the anatomical category and the current surgical stage, a time-varying setting is performed on the anisotropic parameters. The anisotropic parameters include distance weights in at least three principal directions. The time-varying setting dynamically adjusts the anisotropic parameters as the surgical stage progresses, and all distance weights are greater than 0. S3. Perform anisotropic Euclidean distance transformation on the binary mask based on the anisotropic parameters to obtain anisotropic distance fields; S4. Extract multi-level safety equidistant surfaces from the anisotropic distance field based on multi-level distance thresholds. The multi-level distance thresholds include at least a first-level close-to-boundary threshold, a second-level conventional warning boundary threshold, and a third-level long-term warning boundary threshold. The multi-level safety equidistant surfaces are represented by triangular meshes and determined based on the first-level safety distance, the second-level safety distance, and the third-level safety distance. S5. Encapsulate the multi-level security isometric surface and complete metadata into a security boundary data packet. The complete metadata includes at least key anatomical structure categories, anisotropic parameters, distance threshold levels, generation timestamps, volume statistics, and surgical stage labels. The multi-level security isometric surface and the complete metadata are packaged together. S6. The security boundary data packet is traceably exported based on multiple standard 3D model file formats, including at least one of STL, OBJ, PLY, and NIFTI. The traceable export includes retaining all parameter information of the generation process to support retrospective auditing. The security boundary data can be overlaid on the preoperative planning view.

[0006] In any embodiment of this application, the anisotropy parameter is adjusted over time with the surgical stage to reflect the different requirements for the safety boundary range at different surgical stages.

[0007] In any embodiment of this application, the anisotropic Euclidean distance transformation is performed based on the following formula: , Where w_i is the distance weight in the i-th principal direction.

[0008] In any embodiment of this application, the multi-level distance threshold includes at least three levels: Level 1 close proximity, Level 2 regular alert, and Level 3 long-term alert.

[0009] In any embodiment of this application, the complete metadata includes at least the following fields: anatomical category, anisotropy parameter, distance threshold level, timestamp, volume statistics, and surgical stage label.

[0010] In conjunction with any embodiment of this application, the method supports traceable export of multiple standard 3D model file formats such as STL / OBJ / PLY / NIFTI, and supports linkage with preoperative surgical planning systems.

[0011] In a second aspect, a surgical safety boundary generation device based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation is provided, comprising: a mask and category acquisition unit, an anisotropic parameter time-varying setting unit, an EDT calculation unit, a safety surface extraction unit, a metadata encapsulation unit, and a multi-format export unit, which respectively execute steps S1 to S6 of the method described in the first aspect.

[0012] Thirdly, an electronic device is provided, including a processor and a storage unit, the storage unit being used to store computer program code, the code containing computer instructions, wherein when the processor executes these instructions, the electronic device performs the method described in the first aspect.

[0013] Fourthly, a computer-readable storage medium is provided, wherein a computer program is stored therein, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method described in the first aspect.

[0014] Fifthly, a computer program product is provided, the computer program product comprising a computer program or instructions that, when the computer program or instructions are run on a computer, cause the computer to perform the method described in the first aspect.

[0015] It should be understood that the above general descriptions and subsequent specific descriptions are for illustrative and explanatory purposes only and do not impose any limitations on this application.

[0016] In this application, compared with the prior art, the technical problem to be solved by this application is: to realize anisotropic distance transformation based on key anatomical structure categories in the generation of surgical safety boundaries, to realize time-varying setting of anisotropic parameters based on surgical stage, to provide multi-level safety boundaries to support refined surgical planning, to provide complete metadata encapsulation to support downstream interpretability, to provide traceable export of multiple standard 3D model file formats, and to provide bidirectional linkage with the preoperative surgical planning system. The technical means employed in this application and their non-obvious aspects compared to existing technologies include: First, this application is based on anisotropic distance transformation of key anatomical structure categories, a design not seen in existing methods using isotropic distance transformation. This anisotropic transformation addresses the differences in clinical risk of different types of anatomical structures in different directions (e.g., higher vertical risk for vascular structures); Second, this application is based on time-varying anisotropic parameters during the surgical phase, a design not seen in existing methods using fixed parameters. This time-varying setting expands the safety margin during the fine dissection phase and moderately reduces it during the coarse dissection phase, reflecting the differentiated safety margin requirements of different surgical phases; Third, this application provides a three-level safety distance: Level 1 close proximity, Level 2 routine alert, and Level 3 long-term alert. This design is not seen in existing methods that only provide a single threshold. This multi-level mechanism supports refined surgical planning; Fourth, this application is based on Marching... The Cubes algorithm achieves robust and secure isometric surface extraction, a design not seen in existing methods involving simple morphological operations. Fifth, the secure boundary output of this application encapsulates complete metadata (anatomical category + anisotropy parameters + distance threshold level + timestamp + volume statistics + surgical stage label), a design not seen in existing methods that only output geometric data. This metadata encapsulation supports downstream interpretability. Sixth, this application supports traceable export of four standard 3D model file formats: STL / OBJ / PLY / NIFTI, a design not seen in existing methods with a single format. This multi-format support provides flexibility for downstream applications. Seventh, this application integrates secure boundary generation with the preoperative surgical planning system in a bidirectional manner, a design not seen in existing methods that operate separately. This bidirectional integration eliminates the information gap between preoperative planning and secure boundaries. The related devices described in this application include devices, electronic devices, computer-readable storage media, and computer program products. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be explained below.

[0018] Figure 1 This is a flowchart illustrating a surgical safety boundary generation method based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation, provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of a surgical safety boundary generation device based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation, provided in an embodiment of this application.

[0020] Figure 3 This is a schematic diagram of the hardware architecture of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] To enable those skilled in the art to more fully understand the technical solution of this application, the technical solution of this application will be explained in detail and clearly with reference to the accompanying drawings.

[0022] Please see Figure 1 , Figure 1 This is a flowchart illustrating a surgical safety boundary generation method based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation, provided in an embodiment of this application. The method includes steps S1 to S6.

[0023] S1. Mask and Category Acquisition: Acquire the binary mask and corresponding anatomical category of key anatomical structures.

[0024] In this embodiment, the binary mask can be obtained based on preoperative CT or MRI images using a deep learning segmentation network (such as 3D U-Net, V-Net, or nnU-Net). The anatomical category includes at least one of the following: vascular, neural, and tubular.

[0025] S2, Anisotropic parameter time-varying setting: Perform time-varying setting on anisotropic parameters based on the anatomical category and the current surgical stage.

[0026] In this embodiment, the anisotropy parameters include distance weights in at least three principal directions. When the key anatomical structure is a blood vessel, w=0.5 along the vessel's direction and w=1.0 vertically, reflecting the clinical understanding that vascular injury is mainly caused by the vertical approach of instruments. When it is a nerve, w=1.0 in each direction, reflecting that the risk of nerve injury is consistent in all directions. When it is a canal, w=0.5 along the canal's direction and w=1.0 vertically. The parameters are also dynamically adjusted according to the surgical stage. In the fine dissection stage, the distance weights are increased overall to expand the safety margin, and in the coarse dissection stage, they are appropriately reduced to improve surgical efficiency.

[0027] S3. Anisotropic EDT calculation: Perform anisotropic Euclidean distance transformation on the binary mask based on the anisotropic parameters.

[0028] In this embodiment, the anisotropic Euclidean distance transformation is performed based on the following formula: , Where D(p) is the anisotropic distance value at pixel p, M is the set of all pixels in the mask belonging to the key anatomical structure, and w_i is the distance weight in the i-th principal direction. All distance weights are positive real numbers.

[0029] S4. Multi-level safety surface extraction: Extract multi-level safety equidistant surfaces from the anisotropic distance field based on multi-level distance thresholds.

[0030] In this embodiment, the multi-level distance threshold includes at least three levels: the first level of safety distance corresponds to the immediate boundary (range 1 to 3 mm); the second level of safety distance corresponds to the regular warning boundary (3 to 8 mm); and the third level of safety distance corresponds to the long-term warning boundary (8 to 15 mm). The extraction of the safety equidistant surface is performed based on the Marching Cubes algorithm and represented as a triangular mesh (vertex + facet) or voxels (sparse mesh / octree).

[0031] S5. Complete metadata encapsulation: The multi-level security equidistant surface and complete metadata are encapsulated into a security boundary data packet.

[0032] In this embodiment, the complete metadata includes at least the following fields: key anatomical structure category, anisotropy parameters, distance threshold levels, generation timestamp, volume per level statistics, surgical phase label, original binary mask source, and generation method version. The metadata encapsulation is based on structured data formats such as JSON / XML / YAML and is packaged together with the multi-level secure isometric surface and the complete metadata.

[0033] S6. Multi-format traceable export: The security boundary data packet can be traceably exported based on multiple standard 3D model file formats.

[0034] In this embodiment, the various standard 3D model file formats include, but are not limited to: STL format (for viewing in general 3D modeling software); OBJ format (supports visualization of textures and materials); PLY format (includes vertex attributes, used for scientific research analysis); and NIFTI format (for integration into medical imaging systems). The traceable export retains all parameter information of the generation process, supporting retrospective auditing. The method also includes linking safety boundary data with the preoperative surgical planning system. This linking includes overlaying the safety boundary onto the preoperative planning view and adjusting the distance threshold and anisotropy parameters in reverse based on the preoperative planning.

[0035] Application Example 1: In the planning scenario of laparoscopic radical colectomy, the method performs anisotropic EDT based on a 3D mask of the superior mesenteric artery constructed by preoperative CT. During the fine dissection stage, w=0.5 in the lateral direction and w=1.2 in the vertical direction (overall elevation). Three levels of safety boundaries are generated (2mm close proximity / 5mm standard warning / 10mm long-term warning). The complete metadata encapsulation includes the anatomical category "vascular" + anisotropic parameters + volume statistics for each level {2mm: 1.2cc, 5mm: 8.5cc, 10mm: 45cc} + surgical stage "fine dissection" + timestamp. It is exported to the preoperative planning software in STL format to assist in planning the safe range of lymph node dissection.

[0036] Application Example 2: In the planning scenario of laparoscopic total mesorectal resection for rectal cancer, an isotropic EDT (neural class, w=1.0 in each direction) is performed based on a 3D mask of the pelvic autonomic nerves constructed from preoperative MRI. A 3mm safety boundary is generated and exported to the navigation system in NIFTI format to assist the surgeon in operating along the anatomical plane of the membrane. During the preoperative planning stage, the surgeon adjusts the safety boundary from 3mm to 4mm, and the system updates the distance threshold and metadata accordingly.

[0037] Please see Figure 2 , Figure 2 This is a schematic diagram of a surgical safety boundary generation device based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation, provided in an embodiment of this application. The surgical safety boundary generation device 1 includes: The mask and category acquisition unit 11 is used to acquire the binary mask and anatomical category of key anatomical structures.

[0038] The anisotropic parameter time-varying setting unit 12 is used to perform time-varying settings on the anisotropic parameters based on the anatomical category and the current surgical stage.

[0039] EDT calculation unit 13 is used to perform anisotropic Euclidean distance transformation on the binary mask.

[0040] The safety surface extraction unit 14 is used to extract multi-level safety equidistant surfaces from the distance field based on multi-level distance thresholds.

[0041] Metadata encapsulation unit 15 is used to encapsulate multi-level security equidistant surfaces and complete metadata into a security boundary data packet.

[0042] The multi-format export unit 16 is used to traceably export security boundary data packets based on multiple standard 3D model file formats.

[0043] Please see Figure 3 , Figure 3 This is a schematic diagram of the hardware architecture of an electronic device provided in an embodiment of this application. The electronic device 2 includes a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21, the memory 22, the input device 23, and the output device 24 are connected in communication via a bus.

[0044] The method described in this application can be implemented based on hardware, software, or a combination of hardware and software.

Claims

1. A method for generating a surgical safety boundary based on time-varying anisotropic distance transformation and multi-level security tile data encapsulation, characterized in that It includes the following steps: S1. Obtain a binary mask of the key anatomical structure and its corresponding anatomical category. The binary mask represents the occupancy of the key anatomical structure in three-dimensional space, and the anatomical category includes at least one of the following: blood vessel, nerve, and canal. S2. Based on the anatomical category and the current surgical stage, a time-varying setting is performed on the anisotropy parameters. The anisotropy parameters include distance weights in at least three principal directions. The time-varying setting dynamically adjusts the anisotropy parameters as the surgical stage progresses, and all distance weights are greater than 0. S3. Perform anisotropic Euclidean distance transformation on the binary mask based on the anisotropic parameters to obtain anisotropic distance fields; S4. Extract multi-level safety equidistant surfaces from the anisotropic distance field based on multi-level distance thresholds. The multi-level distance thresholds include at least a first-level close-to-boundary threshold, a second-level conventional warning boundary threshold, and a third-level long-term warning boundary threshold. The multi-level safety equidistant surfaces are represented by triangular meshes and determined based on the first-level safety distance, the second-level safety distance, and the third-level safety distance. S5. Encapsulate the multi-level security isometric surface and complete metadata into a security boundary data packet. The complete metadata includes at least key anatomical structure categories, anisotropic parameters, distance threshold levels, generation timestamps, volume statistics, and surgical stage labels. The multi-level security isometric surface and the complete metadata are packaged together. S6. The security boundary data packet is traceably exported based on multiple standard 3D model file formats, including at least one of STL, OBJ, PLY, and NIFTI. The traceable export includes retaining all parameter information of the generation process to support retrospective auditing. The security boundary data can be overlaid on the preoperative planning view.

2. The method of claim 1, wherein The time-varying settings are based on the following rules: when the key anatomical structure is a blood vessel, the distance weight along the direction of the blood vessel is less than the distance weight perpendicular to the direction of the blood vessel; when the key anatomical structure is a nerve, the distance weights in each direction are nearly consistent, reflecting that the risk of nerve injury is consistent in each direction; when the key anatomical structure is a canal, the distance weight along the direction of the canal is less than the distance weight perpendicular to the direction of the canal; the time-varying settings dynamically adjust the anisotropy parameters according to the surgical stage, increasing the overall distance weight in the fine separation stage to expand the safety boundary, and appropriately reducing the distance weight in the coarse dissection stage to improve surgical efficiency.

3. The method of claim 1, wherein The anisotropic Euclidean distance transformation is performed based on the following formula: ， Where D(p) is the anisotropic distance value at pixel p, M is the set of all pixels belonging to key anatomical structures in the binary mask, q is a pixel in the mask, w_i is the distance weight in the i-th principal direction, and p_i and q_i are the coordinates of pixels p and q in the i-th principal direction; all distance weights are positive real numbers; the anisotropic distance transformation, relative to the isotropic distance transformation, can reflect the differences in clinical risk in different directions.

4. The method of claim 1, wherein The multi-level distance threshold includes at least three levels: the first-level safety distance corresponds to the close boundary of the critical anatomical structure, and the value of the first-level safety distance ranges from 1 mm to 3 mm; the second-level safety distance corresponds to the regular warning boundary, and the value of the second-level safety distance ranges from 3 mm to 8 mm; the third-level safety distance corresponds to the long-term warning boundary, and the value of the third-level safety distance ranges from 8 mm to 15 mm; the extraction of the safety equidistant surface is performed based on the Marching Cubes algorithm; the safety equidistant surface is represented by a triangular mesh or voxels, the triangular mesh representation includes a vertex coordinate list and a triangular facet index list, and the voxel representation is stored using a sparse mesh or octree structure to reduce storage overhead.

5. The method according to claim 1, characterized in that... The complete metadata includes at least the following fields: key anatomical structure category, anisotropy parameter, distance threshold level, generation timestamp, volume statistics, surgical stage label, source of original binary mask, and generation method version; the metadata encapsulation is based on a structured data format, which includes at least one of JSON, XML, and YAML, and the multi-level secure isometric surface is packaged together with the complete metadata.

6. The method according to claim 1, characterized in that... The traceable export includes at least the following functions: supporting the export of triangular meshes based on STL format for viewing in general 3D modeling software; supporting the export of triangular meshes based on OBJ format for supporting visualization of textures and materials in scenes; supporting the export of triangular meshes containing vertex attributes based on PLY format for scientific research analysis; supporting the export of voxel representations based on NIFTI format for integration into medical imaging systems; the method also includes linking safety boundary data with a preoperative surgical planning system, the linking including overlaying the safety boundary onto the preoperative planning view, and adjusting the distance threshold and the anisotropy parameters in reverse based on the preoperative planning; the safety boundary data is also used for at least one of the following downstream applications: surgical robot motion constraints, surgical risk quantification assessment, surgical teaching and training, and surgical quality control review.

7. A surgical safety boundary generation device based on time-varying anisotropic distance transformation and multi-level safety surface metadata encapsulation, characterized in that... ,include: A mask and category acquisition unit, configured to perform step S1 as described in claim 1; An anisotropic parameter time-varying setting unit is used to perform step S2 as described in claim 1; An anisotropic distance transformation unit is used to perform step S3 as described in claim 1; A multi-level safety surface extraction unit is used to perform step S4 as described in claim 1; Metadata encapsulation unit, used to perform step S5 as described in claim 1; A multi-format traceable export unit is used to perform step S6 as described in claim 1.

8. An electronic device, characterized in that... The electronic device includes: a processor and a storage unit for storing computer program code containing computer instructions, wherein when the processor executes these instructions, the electronic device performs the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that... The computer-readable storage medium stores a computer program, the computer program containing program instructions that, when executed by a processor, cause the processor to perform the method described in any one of claims 1 to 6.

10. A computer program product, characterized in that... The computer program product comprises a computer program or instructions that, when executed on a computer, cause the computer to perform the method described in any one of claims 1 to 6.

Images