文献速递技术与安全CTMR

大语言模型生成放射报告的易读性改善与翻译不稳定性

Improved Readability and Translational Instability in LLM-Generated Radiology Reports.
PubMed 文献 · Acad Radiol · 2026-08-06

本研究回顾性评估三种大语言模型将320份放射报告转换为患者友好语言的翻译准确性、错误率及不稳定性。采用两阶段提示工程策略,由两名资深放射科医生评价医学准确性,16名非医学参与者评价易读性。结果显示优化提示可降低输出变异并提高翻译准确性,DeepSeek-R1表现最佳,但模型仍存在信息遗漏,目前仅限人工监督下使用。

要点

  • 所有测试模型均存在输出不稳定性、信息遗漏及生成偏保守的宽泛临床建议。
  • 优化结构化提示可显著降低模型输出变异,DeepSeek-R1与ChatGPT-4.0改善最突出。
  • 大语言模型当前仅应作为人工监督的辅助工具,不能独立用于临床。

核心结论

  • 大语言模型可显著改善放射报告的易读性,DeepSeek-R1表现最佳。
  • 所有测试模型固有存在输出不稳定与信息遗漏问题。
  • 优化结构化提示可大幅降低模型输出变异并提高医学文本翻译准确性。
  • 大语言模型目前应严格限于人工监督下的辅助工具,不宜作为独立临床解决方案。

研究设计与样本

  • 回顾性研究,评估320份去标识化放射报告。
  • 三种大语言模型采用两阶段(基线与优化)提示工程策略处理报告。
  • 两名资深放射科医生评价医学准确性、完整性与建议适宜性。
  • 16名非医学参与者按年龄与教育程度分层评价易读性。

诊断标准与关键数值

  • 大语言模型整体显著提升放射报告易读性(P < 0.05)。
  • DeepSeek-R1在易读性方面取得最佳表现。
  • 优化结构化提示显著降低模型输出方差并提高翻译准确性。
  • 患者对报告的自述理解程度受人口学特征影响。

临床应用与阅片要点

  • 优化结构化提示应纳入放射报告翻译工作流程以减少输出变异。
  • 模型生成的报告中易出现信息遗漏,阅片医生核签时需重点比对关键发现是否完整。
  • 模型倾向生成保守宽泛的临床建议,不能替代放射科医生的具体随访建议。
  • 不同年龄与教育程度患者对报告理解存在差异,发放患者版报告时需考虑该因素。

局限性

  • 回顾性研究设计。
  • 非医学参与者样本量较小(16名)。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

BACKGROUND: Large language models (LLMs) show promise for converting complex radiology reports into patient-centric language, but inherent output instability may limit clinical application.

OBJECTIVES: To quantitatively assess the translational accuracy, error rates, and instability of various LLMs when generating patient-centric radiology reports, and evaluate demographic influences on report readability.

MATERIALS AND METHODS: This retrospective study evaluated 320 de-identified radiology reports processed by three LLMs using a two-stage (baseline and optimized) prompt engineering strategy. Two senior radiologists evaluated medical accuracy, completeness, and recommendation suitability. Readability was evaluated by 16 non-medical participants stratified by age and education.

RESULTS: Professional radiological evaluation revealed that all tested models exhibited inherent instability, omitted information, and tended to generate risk-averse, generalized clinical recommendations. To address these limitations, optimized structured prompts significantly reduced model output variance and improved translational accuracy, with particularly prominent effects observed in DeepSeek-R1 and ChatGPT-4.0. Overall, large language models significantly enhanced the readability of radiology reports (P < 0.05), with DeepSeek-R1 achieving the best performance. However, patients' self-reported comprehension of the reports was affected by demographic characteristics.

CONCLUSION: Large language models can effectively improve the readability of radiology reports, yet all such models inherently suffer from output instability and information omission. Optimized structured prompting can substantially reduce the variability of model outputs and improve the accuracy of medical text translation. Nevertheless, LLMs should currently be strictly confined to human-supervised auxiliary tools rather than applied as standalone clinical solutions.

来源:Acad Radiol,查看原文。版权归原作者所有。

引用

  1. Mao, Wang, Wang, et al. Improved Readability and Translational Instability in LLM-Generated Radiology Reports. Acad Radiol. 2026. PMID: 42562714. doi:10.1016/j.acra.2026.07.040

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。