Debiasing Counterfactuals In the Presence of Spurious Correlations

TL;DR

This paper presents an end-to-end framework combining debiasing classifiers and counterfactual image generation to reduce reliance on spurious correlations in medical imaging, improving model generalization.

Contribution

It introduces a novel integrated training approach and a new metric, SCLS, to mitigate and quantify spurious correlation reliance in deep learning models.

Findings

Debiasing improves generalization across datasets.

Models focus on disease pathology rather than artifacts.

The framework effectively ignores spurious correlations.

Abstract

Deep learning models can perform well in complex medical imaging classification tasks, even when basing their conclusions on spurious correlations (i.e. confounders), should they be prevalent in the training dataset, rather than on the causal image markers of interest. This would thereby limit their ability to generalize across the population. Explainability based on counterfactual image generation can be used to expose the confounders but does not provide a strategy to mitigate the bias. In this work, we introduce the first end-to-end training framework that integrates both (i) popular debiasing classifiers (e.g. distributionally robust optimization (DRO)) to avoid latching onto the spurious correlations and (ii) counterfactual image generation to unveil generalizable imaging markers of relevance to the task. Additionally, we propose a novel metric, Spurious Correlation Latching Score (SCLS), to quantify the extent of the classifier reliance on the spurious correlation as exposed by the counterfactual images. Through comprehensive experiments on two public datasets (with the simulated and real visual artifacts), we demonstrate that the debiasing method: (i) learns generalizable markers across the population, and (ii) successfully ignores spurious correlations and focuses on the underlying disease pathology.