Improving Performance, Robustness, and Fairness of Radiographic AI Models with Finely-Controllable Synthetic Data

TL;DR

This paper introduces RoentGen-v2, a controllable synthetic data generator for chest radiographs that improves deep learning model performance, robustness, and fairness across diverse patient demographics by using a novel training strategy.

Contribution

We develop RoentGen-v2, the first demographic-conditioned diffusion model for clinically plausible chest radiograph synthesis, and demonstrate its effectiveness in enhancing model performance and fairness.

Findings

Synthetic pretraining improves classification accuracy by 6.5%.

Fairness gap reduces by 19.3% with synthetic pretraining.

Synthetic data enhances model generalization across institutions.

Abstract

Achieving robust performance and fairness across diverse patient populations remains a challenge in developing clinically deployable deep learning models for diagnostic imaging. Synthetic data generation has emerged as a promising strategy to address limitations in dataset scale and diversity. We introduce RoentGen-v2, a text-to-image diffusion model for chest radiographs that enables fine-grained control over both radiographic findings and patient demographic attributes, including sex, age, and race/ethnicity. RoentGen-v2 is the first model to generate clinically plausible images with demographic conditioning, facilitating the creation of a large, demographically balanced synthetic dataset comprising over 565,000 images. We use this large synthetic dataset to evaluate optimal training pipelines for downstream disease classification models. In contrast to prior work that combines real and synthetic data naively, we propose an improved training strategy that leverages synthetic data for supervised pretraining, followed by fine-tuning on real data. Through extensive evaluation on over 137,000 chest radiographs from five institutions, we demonstrate that synthetic pretraining consistently improves model performance, generalization to out-of-distribution settings, and fairness across demographic subgroups. Across datasets, synthetic pretraining led to a 6.5% accuracy increase in the performance of downstream classification models, compared to a modest 2.7% increase when naively combining real and synthetic data. We observe this performance improvement simultaneously with the reduction of the underdiagnosis fairness gap by 19.3%. These results highlight the potential of synthetic imaging to advance equitable and generalizable medical deep learning under real-world data constraints. We open source our code, trained models, and synthetic dataset at https://github.com/StanfordMIMI/RoentGen-v2 .