中医药信息2025，Vol.42Issue(6)：12-17,6.DOI:10.19656/j.cnki.1002-2406.20250603
大语言模型在中医诊疗场景的应用思路探讨
Exploration of Large Language Models Application in Traditional Chinese Medicine Diagnosis and Treatment Scenarios
摘要
Abstract
Intelligent diagnosis and treatment in Traditional Chinese Medicine(TCM)constitutes a crucial component of TCM modernization and standardization.The semantic understanding and knowledge emergence capabilities of large language models(LLMs)provide significant support for the innovative development of intelligent TCM diagnosis and treatment.This paper first outlines the general process of constructing TCM diagnostic LLMs,including large-scale corpus preparation,model framework construction,instruction fine-tuning,and alignment fine-tuning.To enhance model interpretability and mitigate hallucination issues in LLMs,we explore approaches to improve diagnostic efficacy through knowledge graph augmentation.Furthermore,we prospect the development trends of LLMs in TCM diagnosis and treatment.We posit that research on LLMs in TCM diagnostics will substantially support the innovative advancement of TCM practices,potentially becoming vital tools for clinical practice,education,and scientific research,thereby promoting TCM's digital and intelligent transformation.Research on large models in TCM must thoroughly consider TCM theory and patient needs,focusing on accurate modeling of TCM principles,diagnostic data,and knowledge to meet the practical requirements of both patients and practitioners.关键词
大语言模型/中医诊疗/知识图谱/应用思路/发展趋势Key words
Large language model/Traditional Chinese medicine diagnosis and treatment/Knowledge graph/Application approach/Development trend引用本文复制引用
王欣宇,杨涛,谢佳东,何佳怡,胡孔法..大语言模型在中医诊疗场景的应用思路探讨[J].中医药信息,2025,42(6):12-17,6.基金项目
国家自然科学基金项目(82074580,82174276) （82074580,82174276）
江苏省重点研发计划项目(BE2022712) （BE2022712）
江苏高校"青蓝工程"资助项目(2024) （2024）
