The Catholic University of Korea

Gastric pathology model reading whole-slide images through a chain of dependent questions, mirroring a pathologist's stepwise reasoning.

Pathology vision foundation model backbone hybridizing state space models with a ViT. 57% higher leave-one-study-out correlation on gene expression.

Prostate cancer detection model for MRI and transrectal ultrasound, trained with patch-level contrastive learning across 4,401 patients.

Breast cancer recurrence risk read from an H&E slide and six routine clinical variables, using a frozen pan-cancer pathology encoder.

Pathology report generation from multi-scale whole-slide images, conditioning a frozen PubMed-pretrained GPT-2 on organ-keyed tag-attended features.