FairGen made skin image synthesis far more demographically fair while keeping accuracy
FairGen: preference-aligned diffusion for demographically equitable medical image synthesis.
Zhimin Li … Tianlong Chen
NPJ digital medicine · 2026
Background
Medical imaging is central to modern diagnostics, and artificial intelligence (artificial intelligence) systems are increasingly used to support image-based analysis by improving efficiency, accuracy, and access to care.
Results
Fairness across demographic groups improved 95.9% for skin images, accuracy maintained
95.9%
skin images
80%
chest radiography
35.2%
brain MRI
More results
However, inequities in healthcare access and differential disease prevalence create severe demographic imbalances in clinical image data.
Artificial intelligence models trained on such imbalanced data risk perpetuating diagnostic bias and widening healthcare disparities.
“
Conclusion
Clinician-facing expert review and external validation on independent cohorts further support that these gains extend beyond standard fidelity metrics and are not confined to the original in-distribution datasets.