多智能体协作的护理问诊教学小程序构建与评价研究

doi:10.3761/j.issn.1672-9234.2026.07.008

中华护理教育 ›› 2026, Vol. 23 ›› Issue (7): 814-821.doi: 10.3761/j.issn.1672-9234.2026.07.008

多智能体协作的护理问诊教学小程序构建与评价研究

王妙丽¹(), 黄重梅², 肖霖³, 陈琦蓉¹, 张浩杰⁴, 谭茗慧¹, 陶宏亮¹, 刘伟¹, 丁金锋¹^,^*()

¹ 中南大学湘雅护理学院长沙市 410013
² 中南大学计算机学院长沙市 410013
³ 宁夏医科大学护理学院银川市 750004
⁴ 南方医科大学护理学院广州市 510515

收稿日期:2026-01-16 出版日期:2026-07-15 发布日期:2026-07-17
通讯作者: *丁金锋，E-mail：jinfeng.ding@csu.edu.cn
作者简介:王妙丽：女，本科（硕士在读），护师，E-mail：1227830455@qq.com
第一联系人：作者贡献声明
研究构思与设计为王妙丽、丁金锋、刘伟、黄重梅、肖霖，模型训练为张浩杰，数据收集与处理为王妙丽、陶宏亮，论文撰写为王妙丽、丁金锋、黄重梅，英文修订为王妙丽，论文修改和审校为丁金锋、黄重梅、陈琦蓉、谭茗慧、刘伟。
基金资助:
本科教育教学改革重大项目“揭榜挂帅”项目(NYJXJBGS-2024015)

Construction and evaluation of a multi-agent collaborative nursing inquiry teaching mini-program

WANG Miaoli¹(), HUANG Chongmei², XIAO Lin³, CHEN Qirong¹, ZHANG Haojie⁴, TAN Minghui¹, TAO Hongliang¹, LIU Wei¹, DING Jinfeng¹^,^*()

¹ Xiangya School of Nursing， Central South University， Changsha 410013， China
² School of Computer Science， Central South University， Changsha 410013， China
³ School of Nursing， Ningxia Medical University， Yinchuan 750004， China
⁴ School of Nursing， Southern Medical University， Guangzhou 510515， China

Received:2026-01-16 Online:2026-07-15 Published:2026-07-17
Contact: *DING Jinfeng，E-mail：jinfeng.ding@csu.edu.cn
Supported by:
Major Undergraduate Education and Teaching Reform Project “Open Competition for Task Assignment”(NYJXJBGS-2024015)

摘要/Abstract

摘要：

目的构建集知识答疑、模拟问诊与反馈评价于一体的智能护理问诊教学小程序，并评估其可用性。方法基于人工智能（artificial intelligence，AI）伙伴和AI导师智能体协作框架，构建包含助教、病人及专家角色的问诊智能体，具体步骤包括：选择DeepSeek-R1-Distill-Qwen-7B、MMed-Llama-3-8B和DeepSeek-V3作为基座模型，构建包含问诊知识、护患问诊对话与问诊评价专项数据集，并采用监督微调和（或）提示工程进行训练，为三类智能体各生成7个备选版本；初步测评后，邀请19名专家基于QUEST评估框架遴选针对各角色的最优版本；于Coze平台完成智能体优化与发布，并在微信平台部署护理问诊教学小程序。邀请20名护理专业学生和5名专家，分别采用整体性评估可用性问卷与启发式评估问卷，对小程序进行可用性评价。结果采用提示词工程进行训练的DeepSeek-V3模型被确定为最优版本。可用性评价显示，护理专业学生对系统质量、信息质量、界面质量及整体质量的评分依次为（1.41±0.42）分、（1.66±0.57）分、（1.75±0.70）分与（1.56±0.46）分；专家在启发式评估问卷各条目上的中位评分均≤1.00分。结论多智能体协作的护理问诊教学小程序具有良好的可用性，可为护理问诊教学提供智能化支持。

关键词: 人工智能, 大语言模型, 护理问诊, 可用性, 小程序

Abstract:

Objective To develop an intelligent nursing inquiry teaching mini-program integrating knowledge-based question answering，simulated clinical inquiry，and feedback evaluation，and to evaluate its usability. Methods Based on a collaborative framework of artificial intelligence（AI） Partner and AI Tutor agents，inquiry agents representing teaching assistant，patient，and expert roles were constructed. The base models DeepSeek-R1-Distill-Qwen-7B，MMed-Llama-3-8B，and DeepSeek-V3 were selected，and specialized datasets covering inquiry knowledge，nurse-patient dialogue，and inquiry evaluation were constructed. The agents were trained using supervised fine-tuning，prompt engineering，or a combination of both，producing seven candidate versions for each agent type. Following preliminary evaluation，experts selected the optimal version of each role based on the QUEST assessment framework. Agent optimization and deployment were performed on the Coze platform，and the mini-program was launched on the WeChat platform. Usability was evaluated by twenty nursing students and five experts using an overall usability questionnaire and a heuristic evaluation，respectively. Results The DeepSeek-V3 model trained with prompt engineering was selected as the optimal version and was applied to develop the nursing inquiry teaching mini-program. Usability evaluation showed that the scores of 20 nursing students for system quality，information quality，interface quality，and overall quality were 1.41 ± 0.42，1.66 ± 0.57，1.75 ± 0.70，and 1.56 ± 0.46，respectively. In the heuristic evaluation conducted by five experts，the median scores for all items were ≤1.00. Conclusion The nursing inquiry teaching mini-program shows favorable usability and offers intelligent support for nursing inquiry education.

Key words: Artificial Intelligence, Large Language Models, Nursing Inquiry, Usability, Mini-Program

王妙丽, 黄重梅, 肖霖, 陈琦蓉, 张浩杰, 谭茗慧, 陶宏亮, 刘伟, 丁金锋. 多智能体协作的护理问诊教学小程序构建与评价研究[J]. 中华护理教育, 2026, 23(7): 814-821.

WANG Miaoli, HUANG Chongmei, XIAO Lin, CHEN Qirong, ZHANG Haojie, TAN Minghui, TAO Hongliang, LIU Wei, DING Jinfeng. Construction and evaluation of a multi-agent collaborative nursing inquiry teaching mini-program[J]. Chinese Journal of Nursing Education, 2026, 23(7): 814-821.

图/表 5

参考文献 19

[1]	Wilson SF, Giddens JF. Health assessment for nursing practice[M]. 6th ed. St Louis, MO: Elsevier Health Sciences, 2020.
[2]	Shanahan M, McDonell K, Reynolds L. Role play with large language models[J]. Nature, 2023, 623(7987):493-498. doi: 10.1038/s41586-023-06647-8
[3]	Holderried F, Stegemann-Philipps C, Herrmann-Werner A, et al. A language model-powered simulated patient with automated feedback for history taking:prospective study[J]. JMIR Med Educ, 2024, 10:e59213. doi: 10.2196/59213
[4]	Yang DY, Ziems C, Held W, et al. Social skill training with large language models[PP/OL]. V1.(2024-04-05). https://arxiv.org/abs/2404.04204.
[5]	Tordjman M, Liu ZL, Yuce M, et al. Comparative benchmarking of the DeepSeek large language model on medical tasks and clinical reasoning[J]. Nat Med, 2025, 31(8):2550-2555. doi: 10.1038/s41591-025-03726-3
[6]	Wu CY, Qiu PC, Liu JX, et al. Towards evaluating and building versatile large language models for medicine[J]. npj Digit Med, 2025, 8:58. doi: 10.1038/s41746-024-01390-4 pmid: 39865143
[7]	DeepSeek-AI, Liu AX, Feng B, et al. DeepSeek-V3 technical report[PP/OL]. V1.(2025-01-22). https://arxiv.org/abs/2412.194371.
[8]	Wang WP, Zhou YC, Fu JX, et al. Evaluating the performance of DeepSeek-R1 and DeepSeek-V3 versus OpenAI models in the Chinese national medical licensing examination:cross-sectional comparative study[J]. JMIR Med Educ, 2025, 11:e73469. doi: 10.2196/73469
[9]	何欢, 杜景榕, 王子文, 等. 虚拟病人护理问诊教学系统的开发与应用[J]. 中华护理教育, 2021, 18(11):972-977.
	He H, Du JR, Wang ZW, et al. Development and application of a virtual patient nursing inquiry teaching system[J]. Chin J Nurs Educ, 2021, 18(11):972-977.
[10]	陈璐, 刘晓英. 护理本科生护理问诊能力评价指标研究[J]. 中国护理管理, 2012, 12(3):50-53.
	Chen L, Liu XY. Evaluation index of nursing interrogation ability for nursing undergraduate students[J]. Chin Nurs Manag, 2012, 12(3):50-53.
[11]	张钦彤, 王昱超, 王鹤羲, 等. 大语言模型微调技术的研究综述[J]. 计算机工程与应用, 2024, 60(17):17-33. doi: 10.3778/j.issn.1002-8331.2312-0035
	Zhang QT, Wang YC, Wang HX, et al. Comprehensive review of large language model fine-tuning[J]. Comput Eng Appl, 2024, 60(17):17-33. doi: 10.3778/j.issn.1002-8331.2312-0035
[12]	王东清, 芦飞, 张炳会, 等. 大语言模型中提示词工程综述[J]. 计算机系统应用, 2025, 34(1):1-10.
	Wang DQ, Lu F, Zhang BH, et al. Survey on prompt engineering in large language model[J]. Comput Syst Appl, 2025, 34(1):1-10.
[13]	Tam TYC, Sivarajkumar S, Kapoor S, et al. A framework for human evaluation of large language models in healthcare derived from literature review[J]. npj Digit Med, 2024, 7:258. doi: 10.1038/s41746-024-01258-7 pmid: 39333376
[14]	Nielsen J, Molich R. Heuristic evaluation of user interfaces[C]// Proceedings of the SIGCHI Conference on Human Factors in Computing Systems Empowering People-CHI’ 90. April 1-5,1990. Seattle,Washington, USA. ACM, 1990:249-256.
[15]	Nielsen J. Usability inspection methods[C]// Conference Companion on Human Factors in Computing Systems-CHI’ 94. April 24-28,1994. Boston,Massachusetts,USA. ACM, 1994:413-414.
[16]	Lewis J. Psychometric evaluation of the PSSUQ using data from five years of usability studies[J]. Int J Human-Comp Interaction, 2002, 14(3):463-488. doi: 10.1207/S15327590IJHC143&4_11
[17]	Sauro J, Lewis JR. 用户体验度量:量化用户体验的统计学方法[M]. 殷文婧,徐沙,杨晨燕,等,译. 北京: 机械工业出版社, 2014.
	Sauro J, Lewis JR. Quantifying the user experience: practical statistics for user research[M]. Yin WJ,Xu S,Yang CY,et al,trans. Beijing: China Machine Press, 2014.
[18]	Kolb AY. The Kolb learning style inventory-version 3.1 2005 technical specifications[R]. Boston,MA:Hay Resource Direct, 2005:166-171.
[19]	方蘅英, 曲峰蕾, 程龙, 等. 急性左心衰竭临床护理决策思维人工智能训练系统的开发及在新入职护士中的应用研究[J]. 中华护理教育, 2025, 22(9):1049-1056.
	Fang HY, Qu FL, Cheng L, et al. Development of an artificial intelligence training system for clinical nursing decision-making thinking in acute left heart failure and its application effect among newly recruited nurses[J]. Chin J Nurs Educ, 2025, 22(9):1049-1056.

多智能体协作的护理问诊教学小程序构建与评价研究

Construction and evaluation of a multi-agent collaborative nursing inquiry teaching mini-program

HTML阅读

PDF下载

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 19

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	高云, 苏茜, 杜志敏, 周英, 唐秋碧, 廖晓艳. 人工智能赋能护理心理学课程思政教学改革与实践[J]. 中华护理教育, 2026, 23(7): 778-783.
[2]	周莉, 孙圣博, 张松灵, 朴美花, 伦汪洋, 周越洋. 新医科背景下妇产科学课程教学创新实践及启示[J]. 中华护理教育, 2026, 23(7): 784-789.
[3]	王卓, 罗羽, 谭静, 洪燕, 郑馨怡, 王宗华. 健康评估智慧课程平台的构建与应用效果研究[J]. 中华护理教育, 2026, 23(7): 822-829.
[4]	任吉, 王克芳, 张玉琪, 唐春艳, 王亚男, 张琳, 牟静静. 情绪在护理本科生人工智能技术感知使用障碍与学习动机间的中介效应分析[J]. 中华护理教育, 2026, 23(7): 836-841.
[5]	靳文韬, 赵娟娟, 曹希, 白杨, 程丽, 郑晶, 李琨. 护理专业学生对课程中人工智能技术的接受度-应用行为差距及成因分析[J]. 中华护理教育, 2026, 23(6): 689-696.
[6]	陈硕贞, 李永涛, 曾芬莲, 罗倩, 陈瑶, 汤炜炜, 吴雪双, 孔令娜. 人工智能素养评估量表的汉化及在护理专业学生中应用的信效度研究[J]. 中华护理教育, 2026, 23(6): 697-704.
[7]	曹采奕, 刘欣园, 吴丽萍, 谢婉, 卢虹, 韩帅, 张玮纯. 人工智能在护理本科教学中应用的范围综述[J]. 中华护理教育, 2026, 23(6): 704-712.
[8]	林丹妮, 华婧, 徐双燕. 自我效能对护理实习生使用人工智能准备度的影响路径分析[J]. 中华护理教育, 2026, 23(5): 594-600.
[9]	徐召召, 韩瑜, 李雪颖, 李颖菊, 尹作娟. 人工智能辅助教学模式下护理教师角色重塑的范围综述[J]. 中华护理教育, 2026, 23(5): 627-633.
[10]	韩舒羽, 尚少梅, 王艳, 金晓燕, 吴雪, 于明明, 张岩, 王翠丽, 周伟娇, 庞冬, 孙静, 侯睿, 董旭, 周宇彤, 王志稳. 护理专业学生创新能力培养范式与实践[J]. 中华护理教育, 2026, 23(4): 422-426.
[11]	仲朝阳, 刘爱梅, 薛晓燕, 王艺讳, 吕坤芳, 王瑛. 人工智能赋能急危重症护理学实训教学的实践研究[J]. 中华护理教育, 2026, 23(4): 427-432.
[12]	王嘉煜, 秦元梅, 杜羿霏, 张琼, 和青森, 杨雯雯. 基于高意识生成式学习的中医护理学课程思政教学创新实践[J]. 中华护理教育, 2026, 23(4): 443-450.
[13]	刘鑫鑫, 由淑萍, 赵学荣, 覃芹丹, 陈凤辉, 颜萍. 护理本科生人工智能准备度的调查研究[J]. 中华护理教育, 2026, 23(4): 457-463.
[14]	彭歆, 王婷, 张秀英, 孙皎, 李昆, 李南如, 蒋明潭. 学业自我效能感在护理专业学生学习动机与生成式人工智能素养间的中介作用分析[J]. 中华护理教育, 2026, 23(3): 322-328.
[15]	文靖, 杨智慧, 闭晓君, 林劲秋, 赵倩倩, 张萍, 张立力. 基于人工智能技术的健康评估教学平台的构建与初步应用研究[J]. 中华护理教育, 2026, 23(2): 133-139.