Abstract:Deep learning models for medical imaging often exhibit overconfidence, creating safety risks in ambiguous diagnostic scenarios. While Conformal Prediction (CP) provides distribution-free statistical guarantees, standard methods such as Regularized Adaptive Prediction Sets (RAPS) optimize for average efficiency and can mask severe failures on difficult inputs.
We propose an Adaptive Lambda Criterion for RAPS that minimizes the worst-case coverage violation across prediction set size strata. On OrganAMNIST (58,850 abdominal CT images, 11 classes), standard size-optimized RAPS converges to near-deterministic behavior with stratified undercoverage on uncertain samples, while our method achieves 95.72 percent global coverage with average set size 1.09 and at least 90 percent coverage across all strata. Cross-domain validation on PathMNIST (107,180 pathology images, 9 classes) confirms generalizability.
Quantitative Grad-CAM analysis (rho = -0.30, p < 1e-22) shows that multi-label predictions correspond to focused attention on anatomically ambiguous regions. These results demonstrate that the proposed method improves reliability while maintaining efficiency, making it suitable for safety-critical medical AI applications.
| Comments: | To appear in IEA/AIE 2026 (Springer LNAI) |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) |
| Cite as: | arXiv:2605.12917 [cs.CV] |
| (or arXiv:2605.12917v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2605.12917 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: One Octadion [view email]
[v1]
Wed, 13 May 2026 02:45:07 UTC (2,295 KB)