Quantum machine learning shows mixed fairness in medical image diagnosis
Responsible Quantum Machine Learning for Medical Imaging: A Fairness Audit Across Sensitive Subgroups
Machine LearningArtificial IntelligenceComputers and Society
Summary
Machine learning can help diagnose medical images, but it may treat different groups of people unfairly. The authors studied classical and quantum machine learning models to see how fair they are across groups like skin types and age. They found quantum models can match overall accuracy but sometimes have unfair differences between groups. Some methods to reduce unfairness lowered accuracy too much, so there’s a trade-off. The results show that quantum models need careful testing to avoid hidden biases.
What this means in practice
- •For medical imaging teams: Improve evaluation of fairness across patient groups by testing classical and quantum models using subgroup-aware metrics on medical images.
- •For quantum computing engineers: Design quantum classifiers with fairness regularization that balance overall accuracy and subgroup performance for sensitive medical data.
Tested on simulated data.
Authors
Giovanni dos Santos Franco, Felipe Mahlow, Ellison Fernando Cardoso, José Adenaldo Santos Bittencourt, Felipe Fanchini
Abstract
Quantum machine learning (QML) is increasingly explored for medical classification, yet its behaviour under subgroup-aware fairness criteria remains understudied. We present an empirical fairness audit of classical and quantum classifiers for medical imaging, assuming QML is not inherently fair. Two scenarios are evaluated: dermatological lesions on DDI (sensitive attribute: Fitzpatrick skin type) and chest X-rays on CheXpert-derived data (sensitive attribute: age group). Using ResNet-18 embeddings, we train SVM, MLP, QSVM and QCNN under a baseline and three mitigation settings: differentiable Equalized Odds (EO) regularisation, Gradient Reversal Learning (GRL), and both combined. Foundation models (PanDerm, EVA-X) serve as external references. Quantum classifiers may achieve competitive aggregate performance while still exhibiting subgroup disparities. Since a smaller subgroup AUC gap may stem from degrading the best group rather than improving the worst, each mitigation is also tested against a pre-specified non-inferiority margin relative to its baseline. Under this criterion, several configurations with the smallest disparity, notably QCNN with GRL on DDI, are not clinically acceptable due to loss of global discrimination. On DDI, QSVM with GRL+EO performed best and approached PanDerm in the performance/fairness trade-off; on CheXpert, QCNN with GRL achieved the highest aggregate performance but increased disparity, and EVA-X remained stronger. Fairness effects are model- and dataset-dependent, so QML requires empirical subgroup-aware evaluation. All quantum models run in classical simulation; we claim no quantum advantage.