Rethinking Bias in Generative Data Augmentation for Medical AI: a Frequency Recalibration Method
This addresses data scarcity and bias in medical AI, which is critical for reliable healthcare applications, though it is an incremental improvement as a post-processing step for existing generative models.
The paper tackles bias in generative data augmentation for medical AI by identifying frequency misalignment between real and synthesized images as a key issue, and proposes the Frequency Recalibration (FreRec) method, which significantly improves downstream medical image classification performance in experiments across brain MRIs, chest X-rays, and fundus images.
Developing Medical AI relies on large datasets and easily suffers from data scarcity. Generative data augmentation (GDA) using AI generative models offers a solution to synthesize realistic medical images. However, the bias in GDA is often underestimated in medical domains, with concerns about the risk of introducing detrimental features generated by AI and harming downstream tasks. This paper identifies the frequency misalignment between real and synthesized images as one of the key factors underlying unreliable GDA and proposes the Frequency Recalibration (FreRec) method to reduce the frequency distributional discrepancy and thus improve GDA. FreRec involves (1) Statistical High-frequency Replacement (SHR) to roughly align high-frequency components and (2) Reconstructive High-frequency Mapping (RHM) to enhance image quality and reconstruct high-frequency details. Extensive experiments were conducted in various medical datasets, including brain MRIs, chest X-rays, and fundus images. The results show that FreRec significantly improves downstream medical image classification performance compared to uncalibrated AI-synthesized samples. FreRec is a standalone post-processing step that is compatible with any generative model and can integrate seamlessly with common medical GDA pipelines.