
浏览全部资源
扫码关注微信
1.上海交通大学医学院附属同仁医院影像科,上海 200336
2.上海交通大学医学院附属同仁医院教学科,上海 200336
3.上海交通大学医学院附属同仁医院院长办公室,上海 200336
4.上海交通大学医学院附属瑞金医院《诊断学理论与实践》编辑部,上海 200025
5.上海交通大学医学院附属瑞金医院放射科,上海 200025
姚伟武,Email:yaoweiwuhuan@163.com
收稿:2024-09-22,
纸质出版:2026-01-20
移动端阅览
钟京谕, 星月, 胡扬帆, 等. 大语言模型辅助住院医师规范化培训课程设计:以胸部X线平片报告教学为例[J]. 中华医学教育探索杂志, 2026, 25(1): 30-36.
Zhong Jingyu, Xing Yue, Hu Yangfan, et al. Large language models in aiding course design for standardized residency training: a case study on teaching the interpretation of chest X-ray plain film reports[J]. Chinese Journal of Medical Education Research, 2026, 25(1): 30-36.
钟京谕, 星月, 胡扬帆, 等. 大语言模型辅助住院医师规范化培训课程设计:以胸部X线平片报告教学为例[J]. 中华医学教育探索杂志, 2026, 25(1): 30-36. DOI: 10.3760/cma.j.cn116021-20240922-01963.
Zhong Jingyu, Xing Yue, Hu Yangfan, et al. Large language models in aiding course design for standardized residency training: a case study on teaching the interpretation of chest X-ray plain film reports[J]. Chinese Journal of Medical Education Research, 2026, 25(1): 30-36. DOI: 10.3760/cma.j.cn116021-20240922-01963.
目的
2
评估大语言模型(large language model,LLM)辅助住院医师规范化培训课程设计的能力和风险。
方法
2
采用个案研究的方法,以胸部X线平片报告教学为例模拟了医学课程设计过程。构建了简体中文提示词要求文心一言和ChatGPT生成课程设计内容,客观记录了两者的回复内容和用时,主观判定了其辅助课程设计的能力和风险。
结果
2
文心一言和ChatGPT单次回复字数的中位数(范围)分别为442(93~653)字和543(117~843)字,回复总字数分别为4 469字和5 362字,对话分别耗时13 min 38 s和12 min 25 s。两者均能辅助课程设计,具备快速编写教案、生成教学材料的能力,但创造性总结的能力较为欠缺,存在虚构内容的风险。
结论
2
大语言模型尚不具备独立设计住院医师规范化培训课程的能力,其生成的内容在用于教学实践前必须经过专家优化和确认后才能用于教学实践,但LLM依然有望成为医学课程设计的助手,甚至撬动医学教育的变革。
Objective
2
To evaluate the ability and risk of using large language models to aid the course design for standardized residency training.
Methods
2
This study employed a case study approach to simulate the process of medical course design
using teaching the interpretation of chest X-ray plain film reports as an example. The prompt words were constructed in simplified Chinese for both ERNIE Bot and ChatGPT to conduct course design. Their responses and response times were objectively recorded
and their ability and risk in aiding course design were subjectively determined.
Results
2
The median (range) of the number of words replied by ERNIE Bot and ChatGPT in one session was 442 (93-653) and 543 (117-843)
the total number of words replied was 4 469 and 5 362
and the response time was 13 min 38 s and 12 min 25 s
respectively. Both large language models were able to aid the course design. They have the ability in quickly writing teaching plans and generating teaching materials. However
they showed limited capability for creative synthesis and carried a risk of generating fabricated information.
Conclusions
2
The large language models do not have the ability to independently design the courses for standardized residency training. The generated information should be optimized and confirmed by experts before being applied in teaching practice. However
large language models hold promise as assistants in medical course design and could potentially drive transformative changes in medical education.
Khurana D , Koli A , Khatter K , et al . Natural language processing: state of the art, current trends and challenges [J ] . Multimed Tools Appl , 2023 , 82 ( 3 ): 3713 - 3744 . DOI: 10.1007/s11042-022-13428-4 http://dx.doi.org/10.1007/s11042-022-13428-4 .
Thirunavukarasu AJ , Ting DSJ , Elangovan K , et al . Large language models in medicine [J ] . Nat Med , 2023 , 29 ( 8 ): 1930 - 1940 . DOI: 10.1038/s41591-023-02448-8 http://dx.doi.org/10.1038/s41591-023-02448-8 .
朱永新 , 杨帆 . ChatGPT/生成式人工智能与教育创新: 机遇、挑战以及未来 [J ] . 华东师范大学学报(教育科学版) , 2023 , 41 ( 7 ): 1 - 14 . DOI: 10.16382/j.cnki.1000-5560.2023.07.001 http://dx.doi.org/10.16382/j.cnki.1000-5560.2023.07.001 .
Kung TH , Cheatham M , Medenilla A , et al . Performance of ChatGPT on USMLE: potential for AI-assisted medical education using large language models [J ] . PLoS Digit Health , 2023 , 2 ( 2 ): e0000198 . DOI: 10.1371/journal.pdig.0000198 http://dx.doi.org/10.1371/journal.pdig.0000198 .
Stoehr F , Kämpgen B , Müller L , et al . Natural language processing for automatic evaluation of free-text answers: a feasibility study based on the European Diploma in Radiology examination [J ] . Insights Imaging , 2023 , 14 ( 1 ): 150 . DOI: 10.1186/s13244-023-01507-5 http://dx.doi.org/10.1186/s13244-023-01507-5 .
贺樑 , 应振宇 , 王英英 , 等 . 教育中的ChatGPT: 教学能力诊断研究 [J ] . 华东师范大学学报(教育科学版) , 2023 , 41 ( 7 ): 162- 176. DOI: 10.16382/j.cnki.1000-5560.2023.07.015 http://dx.doi.org/10.16382/j.cnki.1000-5560.2023.07.015 .
百度 . 文心一言 [EB/OL ] . ( 2023-08-31 ) [ 2023-09-26 ] . https://yiyan.baidu.com https://yiyan.baidu.com .
OpenAI . ChatGPT [EB/OL ] . ( 2022-11-30 ) [ 2023-09-26 ] . https://chat.openai.com https://chat.openai.com .
中国医师协会 . 中国医师协会关于印发住培内容与标准、基地标准(2022年版)的通知 [EB/OL ] . ( 2022-08-05 ) [ 2023-09-01 ] . https://www.ccgme-cmda.cn/news/15117/1/article https://www.ccgme-cmda.cn/news/15117/1/article .
张华平 , 李林翰 , 李春锦 . ChatGPT中文性能测评与风险应对 [J ] . 数据分析与知识发现 , 2023 , 7 ( 3 ): 16 - 25 .
Almeida LC , Farina EMJM , Kuriki PEA , et al . Performance of ChatGPT on the Brazilian radiology and diagnostic imaging and mammography board examinations [J ] . Radiol Artif Intell , 2024 , 6 ( 1 ): e230103 . DOI: 10.1148/ryai.230103 http://dx.doi.org/10.1148/ryai.230103 .
倪闽景 . 从学习进化的视角看ChatGPT/生成式人工智能对学习的影响 [J ] . 华东师范大学学报(教育科学版) , 2023 , 41 ( 7 ): 151 - 161 . DOI: 10.16382/j.cnki.1000-5560.2023.07.014 http://dx.doi.org/10.16382/j.cnki.1000-5560.2023.07.014 .
李袁爽 , 张子航 , 王运武 . 基于ChatGPT的人机共教、共育和共创模式 [J ] . 中国医学教育技术 , 2023 , 37 ( 4 ): 383 - 389, 396 . DOI: 10.13566/j.cnki.cmet.cn61-1317/g4.202304002 http://dx.doi.org/10.13566/j.cnki.cmet.cn61-1317/g4.202304002 .
王斌 , 田士来 , 王小成 , 等 . 以ChatGPT为代表的人工智能技术在医学教育中的机遇与挑战 [J ] . 医学教育研究与实践 , 2023 , 31 ( 4 ): 402 - 405 . DOI: 10.13555/j.cnki.c.m.e.2023.04.002 http://dx.doi.org/10.13555/j.cnki.c.m.e.2023.04.002 .
0
浏览量
1
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621