Error Detection in Emergency Radiology Reports Using a Large Language Model: Multistage Evaluation Study.
Authors
Affiliations (6)
Affiliations (6)
- Department of Radiology, The First Affiliated Hospital of Jinan University, No. 613 Huangpu West Road, Tianhe, Guangzhou, Guangdong, 510630, China, 86 15217921427.
- School of Biomedical Engineering, Southern Medical University, Guangzhou, Guangdong, China.
- Department of Radiology, The Affiliated Hospital of Guangdong Medical University, Zhanjiang, Guangdong, China.
- Department of Radiology, Guangzhou Women and Children's Medical Center, Guangzhou Medical University, Guangzhou, Guangdong, China.
- Department of Radiology, Nanfang Hospital, Southern Medical University, Guangzhou, Guangdong, China.
- Department of Radiology, Longhu District People's Hospital of Shantou, Shantou, Guangdong, China.
Abstract
Emergency radiology requires highly accurate reporting under time constraints; yet, increasing workloads raise the risk of errors. While large language models (LLMs) show potential for proofreading in general radiology, their performance in emergency settings and non-English contexts remains unclear. We aim to evaluate the performance of a domain-optimized LLM, DeepSeek-R1, for identifying errors in Chinese emergency radiology reports, with comparison against assessments by board-certified radiologists. We compiled 7435 emergency reports (dataset 1; radiography, computed tomography, and magnetic resonance imaging) collected from November 2024 to April 2025. In stage 1, a total of 5 LLMs were evaluated using 200 reports. The best model, DeepSeek-R1, proceeded to stages 2 and 3, where 0-shot and few-shot learning were tested on a separate set (n=100). Model performance was compared against 12 radiologists. Stage 4 validated real-world utility on 800 verified reports. In subdataset 1, under stress-testing conditions, DeepSeek-R1 achieved a higher error detection rate in the few-shot setting than in the 0-shot setting (84.4% vs 60.9%, P=.003). Its performance exceeded that of radiology residents (84.4% vs 51.6% and 53.1%, respectively; both P<.05) and showed no statistically significant difference compared with senior radiologists and attending radiologists (84.4% vs 68.8%-93.8%, P=.26 to ≥.99). Compared with residents, DeepSeek-R1 detected more critical omissions (100% vs 25% and 50%; both P<.05) and other errors (92% vs 33% and 33%; both P=.02). In dataset 2, collected from independent institutions, DeepSeek-R1 achieved a detection rate of 95% under the few-shot setting. Reading time was shorter than that of human readers (92 vs 109 s). In real-world validation, DeepSeek-R1 identified 117 true reporting errors, yielding a positive predictive value of 56.5%. DeepSeek-R1 holds promise for improving quality control in emergency radiology reports. Its performance and efficiency support its use as an assistive proofreading tool in real-world radiology workflows.