Disparities in Dermatology AI Performance on a Diverse, Curated Clinical Image Set

TL;DR

This study reveals significant biases in dermatology AI models, showing reduced performance on diverse skin tones and rare diseases, and demonstrates that fine-tuning on diverse datasets can improve accuracy and fairness.

Contribution

The paper introduces the DDI dataset, a curated, diverse dermatology image set, and evaluates AI performance disparities, highlighting the need for bias mitigation in dermatology AI.

Findings

AI models perform 27-36% worse on diverse images

Models and dermatologists perform worse on dark skin and rare diseases

Fine-tuning on diverse data improves AI fairness and accuracy

Abstract

Access to dermatological care is a major issue, with an estimated 3 billion people lacking access to care globally. Artificial intelligence (AI) may aid in triaging skin diseases. However, most AI models have not been rigorously assessed on images of diverse skin tones or uncommon diseases. To ascertain potential biases in algorithm performance in this context, we curated the Diverse Dermatology Images (DDI) dataset-the first publicly available, expertly curated, and pathologically confirmed image dataset with diverse skin tones. Using this dataset of 656 images, we show that state-of-the-art dermatology AI models perform substantially worse on DDI, with receiver operator curve area under the curve (ROC-AUC) dropping by 27-36 percent compared to the models' original test results. All the models performed worse on dark skin tones and uncommon diseases, which are represented in the DDI dataset. Additionally, we find that dermatologists, who typically provide visual labels for AI training and test datasets, also perform worse on images of dark skin tones and uncommon diseases compared to ground truth biopsy annotations. Finally, fine-tuning AI models on the well-characterized and diverse DDI images closed the performance gap between light and dark skin tones. Moreover, algorithms fine-tuned on diverse skin tones outperformed dermatologists on identifying malignancy on images of dark skin tones. Our findings identify important weaknesses and biases in dermatology AI that need to be addressed to ensure reliable application to diverse patients and diseases.