基于状态空间对偶的病理图像分类方法
By using the Mamba2MIL model, which combines state-space duality and multi-instance learning, the problems of low efficiency and error susceptibility in pathological image analysis are solved, achieving efficient whole-slice image classification and improving the accuracy and efficiency of pathological image analysis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIHANG UNIV
- Filing Date
- 2024-08-23
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional pathological image analysis is inefficient and error-prone, especially the challenge of automating the analysis of whole-slice images (WSIs).
By employing the Mamba2MIL model and combining state-space duality (SSD) and multi-instance learning methods, a novel multi-instance learning framework is designed to handle sequential dependencies and independent features in pathological images through sequence enhancement and feature weighted selection. This framework adapts to sequence transformations of different WSI sizes and enhances classification capabilities.
It significantly improves the classification accuracy and analysis efficiency of pathological images, reduces computational complexity, effectively processes long sequence data, and provides more discriminative feature representations.
Smart Images
Figure CN119048825B_ABST