Mammography classification with multi-view deep learning techniques: Investigating graph and transformer-based architectures

Francesco Manigrasso, Rosario Milazzo, Alessandro Sebastian Russo, Fabrizio Lamberti, Fredrik Strand, Andrea Pagnani, Lia Morra

Abstract: Highlights•Extended transformer-based architectures for handling four mammographic views.•Novel cross-view attention module for ipsi-lateral and contra-lateral analysis.•Experimentally compared transformers, convolutional and graph-based architectures.•Transformers with Swin-v2 backbone achieve highest individual performance.•Different architectures have complementary strengths and weaknesses.