欢迎访问新加坡聚知刊出版有限公司官方网站
info@juzhikan.asia
大语言模型在临床诊断中的应用前景与挑战综述
  • ISSN:3029-2816(Online)3029-2808(Print)
  • DOI:10.69979/3029-2808.25.04.057
  • 出版频率:月刊
  • 语言:中文
  • 收录数据库:ISSN:https://portal.issn.org/ 中国知网:https://scholar.cnki.net/journal/search

大语言模型在临床诊断中的应用前景与挑战综述

程睿

四川大学华西临床医学院,四川成都,610041;

摘要:近年来,随着相关技术迅速发展,大语言模型在医疗中的应用不断拓展,而其在临床诊断中的应用与挑战亟待分析总结。本文系统梳理了大语言模型在问诊与病历生成、临床决策支持、影像与报告生成等临床诊断领域的应用探索。随后详细阐述了大语言模型幻觉现象、可解释性、隐私与伦理问题、专科适配性方面存在的挑战与局限,并提出未来解决思路。最后得出结论,大语言模型在临床诊断中展现出显著潜力,其广泛应用仍需在各方面不断完善,最终服务于临床实践。

关键词:大语言模型;人工智能;临床诊断

参考文献

[1]施呈昊,屠馨怡,史佳伟,等.大语言模型临床实践应用范围综述[J].医学信息学杂志,2024,45(9):19-26.

[2]肖仰华,徐一丹.大规模生成式语言模型在医疗领域的应用:机遇与挑战[J].医学信息学杂志,2023,44(9):1-11.

[3]何剑虎,王德健,赵志锐,等.大语言模型在医疗领域的前沿研究与创新应用[J].医学信息学杂志,2024,45(9):10-18.

[4]肖建力,许东舟,王浩,等.医疗领域的大型语言模型综述[J].智能系统学报,2025,20(03):530-547.

[5]CHUA C E, LEE YING CLARA N, FURQAN M S, et al. Integration of customised LLM for discharge summary generation in real-world clinical settings: a pilot study on RUSSELL GPT[J/OL]. The Lancet Regional Health: Western Pacific, 2024, 51: 101211[2025-08-28]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11776078/. DOI:10.1016/j.lanwpc.2024.101211.

[6]HUANG T Y, HSIEH P H, CHANG Y C. Performance Comparison of Junior Residents and ChatGPT in the Objective Structured Clinical Examination (OSCE) for Medical History Taking and Documentation of Medical Records: Development and Usability Study[J/OL]. JMIR medical education, 2024, 10: e59902. DOI:10.2196/59902.

[7]HIROSAWA T, KAWAMURA R, HARADA Y, et al. ChatGPT-Generated Differential Diagnosis Lists for Complex Case-Derived Clinical Vignettes: Diagnostic Accuracy Evaluation[J/OL]. JMIR medical informatics, 2023, 11: e48808. DOI:10.2196/48808.

[8]LIU X, SHI S, ZHANG X, et al. The role of ChatGPT-4o in differential diagnosis and management of vertigo-related disorders[J/OL]. Scientific Reports, 2025, 15(1): 18688. DOI:10.1038/s41598-025-96309-8.

[9]HEWITT K J, WIEST I C, CARRERO Z I, et al. Large language models as a diagnostic support tool in neuropathology[J/OL]. The Journal of Pathology. Clinical Research, 2024, 10(6): e70009. DOI:10.1002/2056-4538.70009.

[10]SERAPIO A, CHAUDHARI G, SAVAGE C, et al. An open-source fine-tuned large language model for radiological impression generation: a multi-reader performance study[J/OL]. BMC medical imaging, 2024, 24(1): 254. DOI:10.1186/s12880-024-01435-w.