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

Samuel Halimi, THEMYR, Arnaud Abreu

Keywords: Chest X-ray, Failure Mode Discovery, Spectral Analysis, Model Robustness, Unsupervised Clustering

TL;DR: We propose a data-centric method using spectral analysis and unsupervised clustering to automatically find model "blind spots" (underperforming data slices) in chest X-rays by detecting subtle, low-level signal variations.

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.

Primary Subject Area: Interpretability and Explainable AI

Secondary Subject Area: Application: Radiology

Registration Requirement: Yes

Visa & Travel: No

Read CFP & Author Instructions: Yes

Originality Policy: Yes

Single-blind & Not Under Review Elsewhere: Yes

LLM Policy: Yes

Submission Number: 233