43 repos across 2 sub-areas
Deep learning approaches for analyzing histopathology and medical images, with a focus on vision transformers, feature extraction, and self-supervised learning methods. This cluster centers on PyTorch-based tools for processing gigapixel pathology slides and extracting interpretable features from medical imagery, with notable work on foundation models like UNI and TITAN that serve as pretrained backbones for downstream pathology tasks. Repositories span implementations of attention-based multiple instance learning, feature aggregation pipelines, and domain-specific vision architectures optimized for the unique computational constraints of whole-slide image analysis.