Beyond Metadata: CAPRA for Hidden Subgroup Analysis under Missing Metadata in Medical Imaging
Once those metadata disappear, clinically critical failure modes can be masked by strong aggregate performance, and many robust-learning methods lose the group structure they rely on. We present CAPRA, a calibrated proxy-axis framework for hidden subgroup analysis under missing metadata. Across fundus, dermoscopy, and chest radiography, CAPRA reveals disparity patterns missed by metadata-only slicing, remains informative under dataset shift, and produces subgroup partitions that align more closely with explicit failure axes than image-only or latent-slice baselines.
- ▪Once those metadata disappear, clinically critical failure modes can be masked by strong aggregate performance, and many robust-learning methods lose the group structure they rely on.
- ▪We present CAPRA, a calibrated proxy-axis framework for hidden subgroup analysis under missing metadata.
- ▪Across fundus, dermoscopy, and chest radiography, CAPRA reveals disparity patterns missed by metadata-only slicing, remains informative under dataset shift, and produces subgroup partitions that align more closely with explicit failure axes
Opening excerpt (first ~120 words) tap to expand
Electrical Engineering and Systems Science > Image and Video Processing arXiv:2607.09102 (eess) [Submitted on 10 Jul 2026] Title:Beyond Metadata: CAPRA for Hidden Subgroup Analysis under Missing Metadata in Medical Imaging Authors:Yawen Li, Yan Li, Zhe Xue, Yingxia Shao, Meiyu Liang, Guanhua Ye View a PDF of the paper titled Beyond Metadata: CAPRA for Hidden Subgroup Analysis under Missing Metadata in Medical Imaging, by Yawen Li and 5 other authors View PDF HTML (experimental) Abstract:Medical imaging models are often deployed without the demographic, acquisition, and quality metadata needed for subgroup auditing.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
