SCGAN-generated skin-tone-balanced extension of the HAM10000 dataset for fairer dermatological AI

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Not clearly reported for clinical outcomes

• Primary Outcome: Generation of 10,000 synthetic skin lesion images with balanced skin tone representation

• Nnt Or Sensitivity: FID scores reported but specific values not provided in abstract

• Confidence Interval: Not provided

Clinical Application

Limited clinical feasibility without validation studies involving real patients and dermatologists Relevant to Australian dermatology given multicultural population, but requires validation with local clinical data and TGA considerations for AI medical devices Dermatological AI development teams and researchers working on diagnostic algorithms

Abstract

This dataset presents a synthetic collection of skin lesion images generated using a tone-conditioned Generative Adversarial Network (GAN) to model skin tone variation in dermatological datasets, motivated by the limited representation of darker skin tones in commonly used benchmarks. The dataset includes 10,000 images, at a resolution of 128×128 pixels, equally divided between medium and dark skin tones, with lesion class labels assigned to follow the class distribution of the original HAM10000 dataset. To quantitatively characterize the generated data, we report Fréchet Inception Distance (FID) scores and a baseline classification experiment comparing models trained on HAM10000 alone versus HAM10000 augmented with the proposed synthetic images. This work provides a controlled synthetic extension intended to support fairness-aware experimentation, data augmentation, and downstream evaluation in dermatological machine learning.

References

1. 1.Kabir, S. (2026). SCGAN-generated skin-tone-balanced extension of the HAM10000 dataset for fairer dermatological AI. PLoS One. https://doi.org/10.1111/exsy.13435