Enhancing Downstream Classification of Breast Abnormalities in Contrast Enhanced Spectral Mammography using a Neighborhood Representation Loss
Medical Imaging 2024: Computer-Aided Diagnosis (SPIE) / 2024 / conference
Bhandary Panambur A, Madhu P, Bayer S, Maier A
Scientific summary
This work studies representation learning for contrast-enhanced spectral mammography to improve downstream abnormality classification.
Abstract
This SPIE paper studies downstream breast abnormality classification in contrast-enhanced spectral mammography using a neighborhood representation loss.
Why it matters
CESM combines morphological and contrast-enhancement information, making robust representation learning important for lesion-level decision support.
Contribution
It explores a representation-learning objective for contrast-enhanced spectral mammography classification.
Method overview
The model is trained so neighboring representations become more clinically useful for downstream abnormality classification.
Key findings
- Published in SPIE Medical Imaging 2024: Computer-Aided Diagnosis.
- Public source metadata verifies the DOI and venue.
- TODO: Add numerical results from the paper when available in the local author manuscript.
Citation
Bhandary Panambur A, Madhu P, Bayer S, Maier A. (2024). Enhancing Downstream Classification of Breast Abnormalities in Contrast Enhanced Spectral Mammography using a Neighborhood Representation Loss. Medical Imaging 2024: Computer-Aided Diagnosis (SPIE).
TODO: Add BibTeX.