Work

Radio-Opaque Artefacts in Digital Mammography: Automatic Detection and Analysis of Downstream Effects

Year: 2025

Type: conference-paper

Abstract: This study investigates the effects of radio-opaque artefacts, such as skin markers, breast implants, and pacemakers, on mammography classification models. After manually anno-tating 22,012 mammograms from the publicly available EM-BED dataset, a robust multi-label artefact detector was de-veloped to identify five distinct artefact types (circular and triangular skin markers, breast implants, support devices and spot compression structures). Subsequent experiments on two clinically relevant tasks - breast density assessment and can-cer screening - revealed that these artefacts can significantly affect model performance, alter classification thresholds, and distort output distributions. These findings underscore the im-portance of accurate automatic artefact detection for developing reliable and robust classification models in digital mam-mography. To facilitate future research our annotations, code, and model predictions are made publicly available1.

Source: IEEE International Symposium on Biomedical Imaging (ISBI)

DOI: 10.1109/isbi60581.2025.10981259

Authors: Amelia Schueppert, Ben Glocker, Melanie Roschewitz

Institutions: Massachusetts Institute of Technology, Imperial College London

Language: English

Topic: AI in cancer detection

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: green

Funders: Royal Academy of Engineering, Imperial College London