Neural fiber bundle clustering segmentation method and system based on dispersion supervoxels
By converting the fiber bundle clustering problem into the clustering problem of diffuse supervoxels, the diffusion weighted undirected graph and geodesic distance are used to solve the problem of high complexity in fiber bundle clustering calculation in the prior art, and efficient fiber bundle clustering and segmentation are achieved, which is suitable for the construction of fiber bundle maps of large sample populations.
Patent Information
- Application Number
- CN202411779568.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-05
- Publication Date
- 2025-04-04
- Estimated Expiration
- 2044-12-05
AI Technical Summary
The existing fiber bundle clustering method has high computational complexity and is difficult to adapt to the construction of fiber bundle maps in large sample populations.
The fiber bundle clustering problem is converted into the clustering problem of diffuse supervoxels. By generating a diffuse weighted undirected graph, calculating the diffuse geodesic distance, and constructing a weighted undirected graph with diffuse supervoxels as vertices, extracting the largest cluster for updating, and realizing clustering and segmentation of fiber bundles.
It reduces the computational complexity, improves the calculation speed, and can handle large-scale fiber curves, provides a feasible solution for fiber bundle map construction in large sample populations, and improves the accuracy and robustness of fiber bundle segmentation.