Self-supervised learning for breast cancer detection: A review

Hugo Figueiras, José Domingues, Nuno Matela, Nuno C. Garcia

Abstract: Highlights•Comprehensive review of SSL across full breast cancer imaging pipeline.•Contrastive learning dominates, yet generative and masked tasks show promise.•Multi-view and multi-resolution SSL boost lesion detection and segmentation.•Provides dataset compendium to accelerate reproducible breast imaging research.

External IDs:dblp:journals/cbm/FigueirasDMG25