A Reasoning-Enabled Vision-Language Foundation Model for Chest X-ray Interpretation

TL;DR

CheXOne is a reasoning-enabled vision-language model for chest X-ray interpretation that generates diagnostic predictions along with explicit, clinically grounded reasoning traces, improving interpretability and performance.

Contribution

The paper introduces CheXOne, a novel model that combines instruction tuning and reinforcement learning to produce explicit reasoning traces in CXR interpretation, enhancing transparency and accuracy.

Findings

CheXOne outperforms existing models on multiple benchmarks.

Clinical study shows reports are comparable or better than resident reports in 55% of cases.

Generated reasoning traces demonstrate high clinical factuality and causal support.

Abstract

Chest X-rays (CXRs) are among the most frequently performed imaging examinations worldwide, yet rising imaging volumes increase radiologist workload and the risk of diagnostic errors. Although artificial intelligence (AI) systems have shown promise for CXR interpretation, most generate only final predictions, without making explicit how visual evidence is translated into radiographic findings and diagnostic predictions. We present CheXOne, a reasoning-enabled vision-language model for CXR interpretation. CheXOne jointly generates diagnostic predictions and explicit, clinically grounded reasoning traces that connect visual evidence, radiographic findings, and these predictions. The model is trained on 14.7 million instruction and reasoning samples curated from 30 public datasets spanning 36 CXR interpretation tasks, using a two-stage framework that combines instruction tuning with reinforcement learning to improve reasoning quality. We evaluate CheXOne in zero-shot settings across visual question answering, report generation, visual grounding and reasoning assessment, covering 17 evaluation settings. CheXOne outperforms existing medical and general-domain foundation models and achieves strong performance on independent public benchmarks. A clinical reader study demonstrates that CheXOne-drafted reports are comparable to or better than resident-written reports in 55% of cases, while effectively addressing clinical indications and enhancing both report writing and CXR interpretation efficiency. Further analyses involving radiologists reveal that the generated reasoning traces show high clinical factuality and provide causal support for the final predictions, offering a plausible explanation for the performance gains. These results suggest that explicit reasoning can improve model performance, interpretability and clinical utility in AI-assisted CXR interpretation.