CorBenchX: Large-Scale Chest X-Ray Error Dataset and Vision-Language Model Benchmark for Report Error Correction
This addresses the problem of automated quality control in clinical radiology reports, though it is incremental as it builds on existing vision-language models and benchmarks.
The authors tackled the lack of a unified benchmark for error detection and correction in radiology reports by introducing CorBenchX, a large-scale dataset and benchmark for chest X-ray reports, where o4-mini achieved the best performance with 50.6% detection accuracy and high correction scores, and they proposed a reinforcement learning framework that improved detection precision by 38.3% and correction by 5.2% over a baseline.
AI-driven models have shown great promise in detecting errors in radiology reports, yet the field lacks a unified benchmark for rigorous evaluation of error detection and further correction. To address this gap, we introduce CorBenchX, a comprehensive suite for automated error detection and correction in chest X-ray reports, designed to advance AI-assisted quality control in clinical practice. We first synthesize a large-scale dataset of 26,326 chest X-ray error reports by injecting clinically common errors via prompting DeepSeek-R1, with each corrupted report paired with its original text, error type, and human-readable description. Leveraging this dataset, we benchmark both open- and closed-source vision-language models,(e.g., InternVL, Qwen-VL, GPT-4o, o4-mini, and Claude-3.7) for error detection and correction under zero-shot prompting. Among these models, o4-mini achieves the best performance, with 50.6 % detection accuracy and correction scores of BLEU 0.853, ROUGE 0.924, BERTScore 0.981, SembScore 0.865, and CheXbertF1 0.954, remaining below clinical-level accuracy, highlighting the challenge of precise report correction. To advance the state of the art, we propose a multi-step reinforcement learning (MSRL) framework that optimizes a multi-objective reward combining format compliance, error-type accuracy, and BLEU similarity. We apply MSRL to QwenVL2.5-7B, the top open-source model in our benchmark, achieving an improvement of 38.3% in single-error detection precision and 5.2% in single-error correction over the zero-shot baseline.