CRCFound
A self-supervised CT foundation model for diagnosis, staging, molecular prediction, and prognosis in colorectal cancer.
Overview
CRCFound is a colorectal cancer CT foundation model designed to turn large volumes of routine unlabeled imaging into transferable clinical intelligence. Instead of depending on task-specific annotation from the start, the model is pretrained on 3D CT scans with self-supervised learning and then adapted to downstream diagnostic and prognostic tasks.
Across staging, molecular prediction, and survival modeling benchmarks, CRCFound shows strong transferability and generalization, especially when labeled data are limited. The project establishes a reusable imaging backbone for colorectal cancer AI rather than a single narrow classifier.