A multimodal slice discovery framework for systematic failure detection and explanation in medical image classification
File(s)
Accepted version
Author(s)
Liu, Yixuan
Bhatia, Kanwal
Fetit, Ahmed
Type
Conference Paper
Abstract
Despite advances in machine learning-based medical image classifiers, the safety and reliability of these systems remain major concerns in practical settings. Existing auditing approaches mainly rely on unimodal features or metadata-based subgroup analyses, which are limited in interpretability and often fail to capture hidden systematic failures. To address these limitations, we introduce the first automated auditing framework that extends slice discovery methods to multi-modal representations specifically for medical applications. Comprehensive experiments were conducted under common failure scenarios using the MIMIC-CXR-JPG dataset, demonstrating the framework’s strong capability in both failure discovery and explanation generation. Our results also show that multimodal information generally allows more comprehensive and effective auditing of classifiers, while unimodal variants beyond image-only inputs exhibit strong potential in scenarios where resources are constrained.
Date Acceptance
2026-01-13
Publisher
IEEE
Copyright Statement
Subject to copyright. This paper is embargoed until publication. Once published the author’s accepted manuscript will be made available under a CC-BY License in accordance with Imperial’s Research Publications Open Access policy (www.imperial.ac.uk/oa-policy).
License URL
Source
2026 IEEE International Symposium on Biomedical Imaging
Publication Status
Accepted
Start Date
2026-04-08
Finish Date
2026-04-11
Coverage Spatial
London, UK