Clinically Aware Synthetic Image Generation for Concept Coverage in Chest X-ray Models

TL;DR

This paper introduces CARPA, a framework for generating anatomically faithful synthetic chest X-ray images with controlled clinical concept modifications to enhance model training and reliability.

Contribution

CARPA is a novel clinically aware and anatomically grounded synthetic image generation method that improves concept coverage and model performance in chest X-ray diagnosis.

Findings

Fine-tuning on CARPA-generated images improves model performance.

Synthetic images maintain high anatomical fidelity and clinical realism.

Expert radiologists confirm the realism and clinical relevance of generated images.

Abstract

Deep learning models for chest X-ray diagnosis are constrained by limited coverage of clinically meaningful concept combinations in publicly available training datasets. While synthetic image generation has been explored to increase data diversity, existing methods rarely enforce clinical or anatomical constraints, limiting utility for improving model reliability. We propose CARPA, a clinically aware and anatomically grounded framework for synthetic chest X-ray generation that applies targeted perturbations to clinical concept vectors while preserving anatomical structure. By producing anatomically faithful synthetic images with controlled concept insertions and deletions, CARPA expands clinically relevant concept coverage. We evaluate CARPA across seven backbone architectures by fine-tuning models on synthetic subsets and testing on a held-out MIMIC-CXR benchmark. Compared to prior concept perturbation approaches, fine-tuning on CARPA-generated images consistently improves precision-recall performance, reduces predictive uncertainty, and improves model calibration. Structural and semantic analyses demonstrate high anatomical fidelity, strong concept alignment, and low semantic uncertainty. Evaluation by two expert radiologists further confirms realism and clinical agreement. Together, these results show that anatomically grounded concept perturbations enable more effective use of synthetic data, improving both performance and reliability of chest X-ray classification models and supporting safer clinical deployment.