Towards Novel Diagnostic Marker Based On Breast Arterial Calcifications

The study explores the integration of advanced computer-aided detection (CAD) systems in clinical practice to improve the detection of Breast Arterial Calcification (BAC) as a potential biomarker for cardiovascular disease (CVD). Conducted in collaboration with Radboud UMC, the research evaluated a novel BAC detector on 500 mammograms, achieving a 70% accuracy rate despite challenges such as heterogeneous data and weak labelling. The study highlights the importance of detecting early BAC stages for accurate CVD assessment and aims to expand the dataset to include more annotated BAC-CVD cases from various institutions to enhance the generalizability of the results.