解放军医学院学报2025，Vol.46Issue(2)：126-133,8.DOI:10.12435/j.issn.2095-5227.24070108
基于医学大模型的智能问诊助手构建研究
Constructing an intelligent consultation assistant system based on medical large language models
摘要
Abstract
Background Traditional medical services depend on face-to-face communication and written medical record documentation,which may affects the efficiency of diagnosis and treatment.The development of AI technology,especially the advancement of large language models,provides the possibilities for transformation in the healthcare industry.Objective To develop an intelligent consultation assistant to improve the quality and efficiency of medical services while offering personalized diagnostic and therapeutic experiences.Methods Based on the comprehensive large language model,digital Homo sapiens technology and 3D visualization technology,the intelligent inquiry system including patient side,server side and doctor side.Results The patient side used the virtual doctor image to communicate with the patient in natural language and collected symptoms and medical history;The server side used the language model of the medical field to analyze the data,and gave a preliminary diagnosis combined with the knowledge base;The doctor side displayed the condition through 3D mannequins to assist doctors to quickly understand and formulate treatment plans.The output scheme continuously fine tuned the model through doctors'confirmation and feedback.This system not only improved the efficiency of information collection and processing but also enhanced the speed of diagnosis through 3D visualization technology.Furthermore,the application of large language models allowed for more precise understanding of conditions,supporting personalized medical services.Conclusion The intelligent consultation assistant not only reduces the workload of doctors,but also provides patients with a more convenient medical experience.关键词
智能问诊助手/数智人/大语言模型/医疗数据可视化/交互体验Key words
intelligent consultation/digital human/large language model/medical data visualization/interaction experience分类
医药卫生引用本文复制引用
庄严,张军雁,卢若谷,于志昊,闻龙,李功,何昆仑..基于医学大模型的智能问诊助手构建研究[J].解放军医学院学报,2025,46(2):126-133,8.基金项目
新一代人工智能国家科技重大专项(2021ZD0140408) （2021ZD0140408）
