Incorporating Rather Than Eliminating: Achieving Fairness for Skin Disease Diagnosis Through Group-Specific Expert

2025-06-21

AI Summary

AI diagnostic models for skin diseases often suffer from population-level bias, leading to a trade-off between fairness and accuracy. Conventional debiasing approaches—by removing correlations with sensitive attributes—risk discarding clinically relevant features, thereby degrading performance. To address this, we propose FairMoE, a novel framework that treats sensitive attributes as modeling guidance rather than nuisance variables. FairMoE employs a hierarchical Mixture-of-Experts (MoE) architecture with dynamic routing, enabling group-specific learning and adaptive assignment of boundary samples. Crucially, it preserves diagnostically discriminative features while explicitly modeling population heterogeneity. Experiments demonstrate that FairMoE achieves fairness metrics—including equal opportunity difference and predictive equality—comparable to or better than state-of-the-art baselines, while significantly improving overall diagnostic accuracy. Thus, FairMoE effectively breaks the conventional fairness–performance trade-off.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Computer Vision: Bias, Fairness & Privacy
• Philosophy and Ethics of AI: Bias, Fairness & Equity

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Economics, Online Markets and Human Computation: Fairness and ethical considerations in crowd work and in human-in-the-loop AI systems

Abstract

AI-based systems have achieved high accuracy in skin disease diagnostics but often exhibit biases across demographic groups, leading to inequitable healthcare outcomes and diminished patient trust. Most existing bias mitigation methods attempt to eliminate the correlation between sensitive attributes and diagnostic prediction, but those methods often degrade performance due to the lost of clinically relevant diagnostic cues. In this work, we propose an alternative approach that incorporates sensitive attributes to achieve fairness. We introduce FairMoE, a framework that employs layer-wise mixture-of-experts modules to serve as group-specific learners. Unlike traditional methods that rigidly assign data based on group labels, FairMoE dynamically routes data to the most suitable expert, making it particularly effective for handling cases near group boundaries. Experimental results show that, unlike previous fairness approaches that reduce performance, FairMoE achieves substantial accuracy improvements while preserving comparable fairness metrics.

Problem

• AI skin diagnosis exhibits biases across demographic groups
• Existing bias mitigation methods degrade diagnostic performance
• FairMoE dynamically routes data to improve accuracy and fairness

Innovation

• Incorporates sensitive attributes for fairness
• Uses group-specific expert learners
• Dynamically routes data to experts