兰州大学学报（医学版）2025，Vol.51Issue(4)：1-7,7.DOI:10.13885/j.issn.1000-2812.2025.04.001
大语言模型驱动的临床决策支持:挑战与实践路径
Large language models for clinical decision support:challenges and implementation pathways
摘要
Abstract
In recent years,the innovative application of large language models has revolutionized traditional clinical decision support systems through their exceptional data integration capabilities and knowledge reason-ing potential.This paper systematically reviewed the technological evolution of large language models in healthcare,analyzing their domain adaptation mechanisms from general language models to medical-special-ized models.While demonstrating advantages across core large language models application scenarios,large language models face fundamental challenges including ensuring clinical accuracy,managing temporal knowl-edge validity and addressing ethical compliance.To address these critical issues,this study proposed a multi-tiered clinical application framework based on cutting-edge research advancements,providing theoretical foundations and actionable pathways for the safe implementation of large language models in clinical decision support.关键词
大语言模型/临床决策支持/人工智能/多模态分析/循证医学/人机协同Key words
large language models/clinical decision support/artificial intelligence/multimodal analysis/evi-dence-based medicine/human-machine collaboration分类
医药卫生引用本文复制引用
周庆国,陈奉贤,李妍..大语言模型驱动的临床决策支持:挑战与实践路径[J].兰州大学学报（医学版）,2025,51(4):1-7,7.基金项目
国家重点研究发展计划资助项目(2024YFE0203800) （2024YFE0203800）
国家自然科学基金资助项目(U22A20261) （U22A20261）
