Hybrid Vision Transformer_GAN Attribute Neutralizer for Mitigating Bias in Chest X_Ray Diagnosis

TL;DR

This paper demonstrates that replacing convolutional encoders with Vision Transformers in an attribute-neutral framework for chest X-ray analysis reduces demographic attribute leakage while maintaining diagnostic accuracy, advancing fairer AI in medical imaging.

Contribution

The study introduces a Vision Transformer-based attribute neutralizer that outperforms convolutional encoders in reducing demographic bias in chest X-ray classifiers.

Findings

Vision Transformer neutralizer reduces sex-recognition AUC to ~0.80.

Diagnostic accuracy remains within 5% of baseline across findings.

Transformer-based approach requires fewer training epochs.

Abstract

Bias in chest X-ray classifiers frequently stems from sex- and age-related shortcuts, leading to systematic underdiagnosis of minority subgroups. Previous pixel-space attribute neutralizers, which rely on convolutional encoders, lessen but do not fully remove this attribute leakage at clinically usable edit strengths. This study evaluates whether substituting the U-Net convolutional encoder with a Vision Transformer backbone in the Attribute-Neutral Framework can reduce demographic attribute leakage while preserving diagnostic accuracy. A data-efficient Image Transformer Small (DeiT-S) neutralizer was trained on the ChestX-ray14 dataset. Its edited images, generated across eleven edit-intensity levels, were evaluated with an independent AI judge for attribute leakage and with a convolutional neural network (ConvNet) for disease prediction. At a moderate edit level (alpha = 0.5), the Vision Transformer (ViT) neutralizer reduces patient sex-recognition area under the curve (AUC) to approximately 0.80, about 10 percentage points below the original framework's convolutional U-Net encoder, despite being trained for only half as many epochs. Meanwhile, macro receiver operating characteristic area under the curve (ROC AUC) across 15 findings stays within five percentage points of the unedited baseline, and the worst-case subgroup AUC remains near 0.70. These results indicate that global self-attention vision models can further suppress attribute leakage without sacrificing clinical utility, suggesting a practical route toward fairer chest X-ray AI.