Predict Patient Self-reported Race from Skin Histological Images

TL;DR

This study demonstrates that deep learning models can predict self-reported race from skin histological images, revealing potential morphological biases and emphasizing the importance of bias mitigation in AI for pathology.

Contribution

It introduces a method to predict race from dermatopathology slides and analyzes morphological features associated with race, highlighting bias risks in AI models.

Findings

High prediction accuracy for White and Black groups (AUC ~0.76-0.80)

Epidermis identified as key predictive region

Bias mitigation strategies impact model performance

Abstract

Artificial Intelligence (AI) has demonstrated success in computational pathology (CPath) for disease detection, biomarker classification, and prognosis prediction. However, its potential to learn unintended demographic biases, particularly those related to social determinants of health, remains understudied. This study investigates whether deep learning models can predict self-reported race from digitized dermatopathology slides and identifies potential morphological shortcuts. Using a multisite dataset with a racially diverse population, we apply an attention-based mechanism to uncover race-associated morphological features. After evaluating three dataset curation strategies to control for confounding factors, the final experiment showed that White and Black demographic groups retained high prediction performance (AUC: 0.799, 0.762), while overall performance dropped to 0.663. Attention analysis revealed the epidermis as a key predictive feature, with significant performance declines when these regions were removed. These findings highlight the need for careful data curation and bias mitigation to ensure equitable AI deployment in pathology. Code available at: https://github.com/sinai-computational-pathology/CPath_SAIF.