一种基于多模态协同感知的脑胶质瘤亚区分割方法及系统
By employing a multimodal collaborative perception-based glioma subregion segmentation method, this approach utilizes the MCM module and HCR Block for feature fusion and enhancement, combined with the GRS module and region consistency correction. This addresses the shortcomings of existing technologies in terms of multimodal information utilization efficiency and subregion recognition capability, achieving more accurate and stable glioma segmentation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANDONG UNIV OF SCI & TECH
- Filing Date
- 2026-05-14
- Publication Date
- 2026-07-17
AI Technical Summary
Existing automatic segmentation methods for gliomas are unable to effectively utilize the complementary relationships between multimodal magnetic resonance images, and cannot simultaneously balance the efficiency of multimodal information utilization, fine-grained subregion recognition ability, and consistency of final output results. In particular, they are not effective in identifying small enhanced tumor targets and the boundaries of diffuse edema.
A glioma subregion segmentation method based on multimodal collaborative perception is adopted. The method uses the MCM module for modal contribution assessment and spatial response modulation, combined with the HCR Block for hierarchical feature extraction and the GRS module for region screening. A three-dimensional convolutional network is used for the encoder-decoder framework, and a region consistency correction strategy is adopted to achieve fine segmentation of gliomas.
It improves the accuracy and stability of glioma subregion segmentation, better identifies smaller, irregularly shaped, and poorly defined enhanced tumor regions, reduces the probability of false-positive segmentation regions, and provides output results with better structural rationality and clinical interpretability.
Smart Images
Figure CN122199982B_ABST