Pathology Foundation Model - Nature Medicine

Multimodal Whole Slide Foundation Model for Pathology

nnMIL: A generalizable multiple instance learning framework for computational pathology

Spatial Expression-Aligned Learning to improve pathology foundation models

A patient-first foundation model for computational pathology. Two-stage training (self-supervised slide encoding + supervised case-level alignment) on 77K+ public whole-slide images across 333 clinical tasks.

A Molecular-Guided Foundation Model with Adaptive Region Modeling for Whole Slide Image Analysis

Pathology whole-slide images (WSIs) pretraining

JEPA-Histo: Joint-Embedding Predictive Architecture for Histopathology Representation Learning

Modular online patch streaming from whole-slide images for training pathology foundation models. No pre-extraction, no disk overhead – patches are sampled, filtered, and augmented on-the-fly.