Revealing Hidden Failure Modes in Chest X-ray Classification via Spectral Domain Analysis

Samuel Halimi, Loic Themyr, Arnaud Abreu
Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, PMLR 315:2682-2710, 2026.

Abstract

Deep learning models for chest X-ray anomaly detection remain vulnerable to subtle distributional shifts (e.g., acquisition technique, patient-related factors, and preprocessing). Traditional error analysis often relies on semantic metadata or model embeddings, which can mask low-level signal variations that degrade performance. In this work, we propose a data-centric framework for automated failure mode discovery using spectral analysis. We project images into the frequency domain and extract a compact profile summarizing the distribution of signal energy across frequency bands. By performing unsupervised clustering on these spectral profiles, we demonstrate that model failures are not randomly distributed, but are strongly concentrated within specific spectral clusters. This method effectively isolates "blind spots", enabling the prediction of model reliability and the discovery of performance-degrading data slices without requiring ground-truth failure annotations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v315-halimi26a, title = {Revealing Hidden Failure Modes in Chest X-ray Classification via Spectral Domain Analysis}, author = {Halimi, Samuel and Themyr, Loic and Abreu, Arnaud}, booktitle = {Proceedings of The 9th International Conference on Medical Imaging with Deep Learning}, pages = {2682--2710}, year = {2026}, editor = {Huo, Yuankai and Gao, Mingchen and Kuo, Chang-Fu and Jin, Yueming and Deng, Ruining}, volume = {315}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v315/main/assets/halimi26a/halimi26a.pdf}, url = {https://proceedings.mlr.press/v315/halimi26a.html}, abstract = {Deep learning models for chest X-ray anomaly detection remain vulnerable to subtle distributional shifts (e.g., acquisition technique, patient-related factors, and preprocessing). Traditional error analysis often relies on semantic metadata or model embeddings, which can mask low-level signal variations that degrade performance. In this work, we propose a data-centric framework for automated failure mode discovery using spectral analysis. We project images into the frequency domain and extract a compact profile summarizing the distribution of signal energy across frequency bands. By performing unsupervised clustering on these spectral profiles, we demonstrate that model failures are not randomly distributed, but are strongly concentrated within specific spectral clusters. This method effectively isolates "blind spots", enabling the prediction of model reliability and the discovery of performance-degrading data slices without requiring ground-truth failure annotations.} }
Endnote
%0 Conference Paper %T Revealing Hidden Failure Modes in Chest X-ray Classification via Spectral Domain Analysis %A Samuel Halimi %A Loic Themyr %A Arnaud Abreu %B Proceedings of The 9th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Yuankai Huo %E Mingchen Gao %E Chang-Fu Kuo %E Yueming Jin %E Ruining Deng %F pmlr-v315-halimi26a %I PMLR %P 2682--2710 %U https://proceedings.mlr.press/v315/halimi26a.html %V 315 %X Deep learning models for chest X-ray anomaly detection remain vulnerable to subtle distributional shifts (e.g., acquisition technique, patient-related factors, and preprocessing). Traditional error analysis often relies on semantic metadata or model embeddings, which can mask low-level signal variations that degrade performance. In this work, we propose a data-centric framework for automated failure mode discovery using spectral analysis. We project images into the frequency domain and extract a compact profile summarizing the distribution of signal energy across frequency bands. By performing unsupervised clustering on these spectral profiles, we demonstrate that model failures are not randomly distributed, but are strongly concentrated within specific spectral clusters. This method effectively isolates "blind spots", enabling the prediction of model reliability and the discovery of performance-degrading data slices without requiring ground-truth failure annotations.
APA
Halimi, S., Themyr, L. & Abreu, A.. (2026). Revealing Hidden Failure Modes in Chest X-ray Classification via Spectral Domain Analysis. Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 315:2682-2710 Available from https://proceedings.mlr.press/v315/halimi26a.html.

Related Material