AI and the World's Hard Problems · Medicine & drug discovery · arXiv · 2026-10-26
AI-guided deep visual proteomics reveals recurrence-associated morpho-molecular signatures within triple-negative breast cancers
AI can localize prognostically relevant regions in digital pathology, but their molecular basis is hard to characterize directly. The authors used recurrence-risk heatmaps to guide physical tissue sampling and mass spectrometry proteomics in 156 triple-negative breast cancer patients, profiling 46 AI-defined high- and low-risk regions from two recurrent cases. A 13-protein composite distinguished highly recurrent from less recurrent patients in an expanded cohort. Combined with the H&E-derived risk score, it raised the C-index from 0.679 to 0.739 and the 5-year time-dependent AUC from 0.675 to 0.770.