MSMV-Swin Framework for Robust Multi-View Breast Cancer Diagnostics
November 2, 2026
The MSMV-Swin framework uses a multi-scale, multi-view Swin Transformer to improve mammography diagnosis robustness against missing views and sparse annotations. Designed as a human-centric decision-support tool, it integrates multi-scale features to assist radiologists in clinical workflows.
HOW THIS AFFECTS YOU
●
researcherThe approach addresses specific clinical failure modes like missing views in medical imaging datasets.
●
healthThis framework provides a more robust decision-support tool for radiologists dealing with incomplete imaging data.