Development and validation of a multimodal AI-agent system for prognosis analysis of bladder urothelial carcinoma - PubMed
5 hours ago
- #Prognosis Analysis
- #Artificial Intelligence
- #Bladder Cancer
- A multimodal AI-agent system named MATCH-Net was developed and validated for bladder urothelial carcinoma (BUC) prognosis analysis.
- The system integrates textual, radiographic, and pathological data, including LLM-standardized pathology reports, CT image segmentation, and histopathological feature extraction from CT scans and whole slide images.
- In multi-center validation with 1185 patients, MATCH-Net achieved robust performance with C-index scores ranging from 0.836 to 0.874.
- It effectively stratified patients into high- and low-risk survival groups and identified candidates potentially responsive to adjuvant chemotherapy.
- The framework quantifies novel interpretable prognostic biomarkers and provides a clinically applicable solution for personalized BUC survival risk stratification.