A team at the German Center for Diabetes Research trained explainable deep‑learning models on gigapixel pancreatic slide images from living donors and identified subtle tissue traits associated with type 2 diabetes. The AI differentiated diabetic from non‑diabetic samples and highlighted alterations in islets and non‑beta endocrine cells tied to impaired insulin secretion. Published in Nature Communications, the study used multiplex immunofluorescence and chromogenic stains to create a large, living‑donor dataset—avoiding post‑mortem artifacts. Explainable AI allowed the researchers to interrogate morphological features driving classification, opening paths for histopathology‑scale biomarker discovery in metabolic disease.