Fairness Evolution in Continual Learning for Medical Imaging
It addresses fairness issues in healthcare AI for medical imaging, but is incremental as it applies existing methods to a new domain-specific problem.
This study tackled the problem of bias evolution in continual learning for medical imaging, finding that while Learning without Forgetting and Pseudo-Label achieve optimal classification performance, Pseudo-Label is less biased.
Deep Learning has advanced significantly in medical applications, aiding disease diagnosis in Chest X-ray images. However, expanding model capabilities with new data remains a challenge, which Continual Learning (CL) aims to address. Previous studies have evaluated CL strategies based on classification performance; however, in sensitive domains such as healthcare, it is crucial to assess performance across socially salient groups to detect potential biases. This study examines how bias evolves across tasks using domain-specific fairness metrics and how different CL strategies impact this evolution. Our results show that Learning without Forgetting and Pseudo-Label achieve optimal classification performance, but Pseudo-Label is less biased.