Construction and clinical teaching application of virtual patient system:based on artificial intelligence LLM technology
周志浩 1宋佳琳 2刘嘉成 3周心悦 4胡汉昆4
扫码查看
点击上方二维码区域，可以放大扫码查看
作者信息
1. 东南大学附属中大医院急诊科(南京 210009)
2. 东南大学医学院(南京 210009)
3. 武汉大学计算机学院(武汉 430072)
4. 武汉大学中南医院药学部(武汉 430071)
折叠
摘要
目的 构建一种基于人工智能大语言模型(large language model,LLM)技术、可用于医学教育的新型虚拟患者(virtual patient,VP)系统,评价该系统在基层医生进修学习全科医学临床思维中的应用效果.方法 选取2021年1月至2024年2月在东南大学附属中大医院进修的基层社区医生为研究对象,随机分为试验组和对照组,分别采用基于LLM的VP系统教学、传统教学方法进行授课,通过临床思维理论知识考核、临床思维能力考核、课程满意度调查评估教学效果,并对结果进行相应的统计学分析.结果 共纳入124名基层社区医生,其中试验组60例、对照组64例,两组在一般基线资料上差异无统计学意义,具有可比性.课程结束后,试验组临床思维理论知识考核成绩显著高于对照组(83.83±3.15 vs.79.92±4.52,P＜0.01),且不及格率显著低于对照组(0.00％vs.9.38％,P＜0.05);试验组在临床思维能力3个维度(批判性、系统性、循证思维)方面教学后分数均显著高于教学前,而对照组仅在批判性思维维度上教学前后差异有统计学意义教学后试验组在系统思维、循证思维方面分数均显著高于对照组(P＜0.05),但在批判性思维上两组分数差异无统计学意义.试验组对授课的总体满意度也显著高于对照组(93.33％vs.85.48％,P＜0.05).结论 基于LLM的VP系统提升了学员对临床思维理论知识的掌握程度,也促进了其临床思维能力的培养,该教学方法可为其他医学教育群体提供新的教学工具和思路.
Abstract
Objective To construct a novel virtual patient(VP)system based on artificial intelligence large language model(LLM)technology for medical education,and to evaluate its effectiveness in improving the clinical thinking skills of primary care physicians in continuing education.Methods Primary care physicians from grassroots community hospitals who were in training at Zhongda Hospital Southeast University from January 2021 to February 2024 were selected as the research subjects and randomly divided into experimental group and control group.The VP system based on LLM and traditional teaching methods were used for instruction.The teaching effects were evaluated through theoretical knowledge assessment,clinical thinking ability assessment,and course satisfaction surveys,with the results subjected to corresponding statistical analysis.Results A total of 124 primary community doctors were included,with 60 in the experimental group and 64 in the control group.No significant statistical differences in the general baseline data between the experimental and control groups.The clinical thinking scores of the experimental group were significantly higher than those of the control group(83.83±3.15 vs.79.92±4.52,P＜0.01),and the failure rate of the experimental group was significantly lower than that of the control group(0.00％vs.9.38％,P＜0.05).The experimental group showed a significant improvement in clinical thinking levels(critical,systematic,and evidence-based thinking)after teaching compared to before,while the control group only showed differences in critical thinking before and after teaching.The overall satisfaction with the course was also higher in the experimental group than in the control group(93.33％vs.85.48％,P＜0.05).Conclusion The VP system based on LLM enhanced the trainees'mastery of clinical thinking theoretical knowledge and also promoted the cultivation of their clinical thinking abilities.The teaching method provides new teaching tools and ideas for other medical education groups.
关键词
人工智能/大语言模型/虚拟患者/医学教育/临床思维
Key words
Artificial intelligence/Large language model/Virtual patient/Medical education/Clinical thinking
维普
万方数据