Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching

Dai Qian; Wei Ke; Zheng Zhi

doi:doi:10.11648/j.sr.20261404.26

Research Article |

| Peer-Reviewed

Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching

Dai Qian, Wei Ke, Zheng Zhi^*

Published in Science Research (Volume 14, Issue 4)

Received: 22 May 2026 Accepted: 15 July 2026 Published: 26 August 2026

Views: Downloads:

Download PDF

Share This Article

Twitter
Linked In
Facebook

Abstract

Modern generative artificial intelligence technologies are fundamentally reshaping the pedagogical paradigm of evidence-based medicine (EBM). This article systematically articulates the mechanisms by which AI augments EBM instruction—encompassing large language models, knowledge graphs, and rule engines—with a specific focus on the "reasoning visualization" pedagogical approach introduced by emerging medical evidence-based AI agents such as DeepRare. In contrast to the functional orientation of general-purpose large language models, DeepRare operationalizes a complete "hypothesis-verification-reflection" reasoning loop, thereby transforming the implicit clinical reasoning processes of conventional instruction into observable, replicable, and critically assessable educational content. Building upon this foundation, the article dissects the concrete pathways of AI integration in EBM teaching across four interdependent dimensions: evidence acquisition and screening, evidence appraisal and the cultivation of critical thinking, evidence synthesis and clinical translation, and teaching evaluation with feedback mechanisms. The inquiry further explores the reconceptualization of teaching models, encompassing competency-adaptive personalized learning trajectories, human-computer collaborative decision-making exercises that promote shared clinical reasoning between learners and intelligent systems, and the "Internet plus evidence-based teaching" paradigm that extends educational impact beyond traditional classroom settings. In confronting practical challenges—including technological dependability, the paucity of empirical validation regarding educational outcomes, ethical and equity considerations, and the digital competency gaps among both instructors and learners—this article advances an integrated four-pronged future agenda comprising technological refinement, curricular redesign, faculty professional development, and deepened empirical research. The ultimate objective is to furnish an operational foundation for the intelligent transformation of EBM education that is congruent with the evolving demands of medical development and talent cultivation in China.

Published in	Science Research (Volume 14, Issue 4)
DOI	10.11648/j.sr.20261404.26
Page(s)	253-260
Creative Commons	This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.
Copyright	Copyright © The Author(s), 2026. Published by Science Publishing Group

Keywords

Evidence-based Medicine, AI, Medical Education, Assistant Teaching, DeepRare

1．引言

循证医学（Evidence-Based Medicine, EBM）教育自提出以来，始终秉持“以最佳研究证据与临床专业技能及患者价值观相结合”的原则。新技术革命时代，现代循证医学教育正面临三重困境：知识大爆炸使得筛选契合“患者价值观”的证据变得愈发困难、临床实践的病因复杂性、对医学生的临床专业技能与决策能力提出了更高的职业素质要求。然而，传统教学模式往往止步于让学生观察到教师“检索-评估-应用”的显性流程，而难以触及更深层次的职业素养。例如，学生无法真正理解在面对复杂临床问题时如何权衡不同证据的价值，也难以习得如何基于最新证据构建“最佳方案”的推理路径。最近，以ChatGPT为代表的生成式人工智能正在深刻改变知识获取与加工的方式。最新《BMJ》调查报告指出，人工智能（Artificial Intelligence，AI）不仅能够扩充证据来源、提高证据获取效率，更能协助医患决策、促进教育科普，展现出创新教学模式的潜力

[1]

，大型语言模型如ChatGPT能够在精准提示下快速生成临床相关案例与多选题，为教育者和学习者提供新型教学工具

[2]

。然而，AI辅助教学在循证医学关键实施“证据获取-评估-整合-应用”各环节的具体作用机制，及其对学生循证思维、推理等的影响仍然不清楚。本研究结合国内外最新文献，聚焦AI技术与循证医学教学深度融合的前沿探索，系统阐述AI在循证医学教学中的应用路径，以期为教学改革提供实践依据。

2．人工智能技术特点

2.1．AI自然语言处理与大型语言模型

以ChatGPT为代表的大型语言模型（LLMs）经过自然语言处理技术的不断发展，能够理解并生成人类自然语言，实现与用户的深度交互

[3]

。研究表明，其核心能力在于能回答连续性问题、质疑用户提出的不合理前提，并拒绝不当请求，从而超越传统聊天机器人的功能，为医学教育提供强大的对话辅助工具

[4]

。最近，国外随机对照研究显示，将ChatGPT融入骨科教学显著提升了学生的测试成绩与满意度，相较于传统教学方法具有明显优势

[5]

。此外，一项将沉浸式虚拟现实与GPT技术相结合应用于临床考试（OSCE）的研究显示，接触最新技术的学生的学习热情和自信心明显增强。研究表明，AI不仅能通过识别学生个体薄弱领域实现个性化学习，还能辅助其理解教学内容、提升临床轮转准备度

[6]

。因此，AI正逐步成为“数字导师”，可按照“技术特性-认知过程-教育价值”三螺旋模式模型实践

[7]

，这将为循证医学教学中的知识强化与技能训练助力个性化和终身学习支持。

2.2．AI知识图谱与规则

基于知识图谱与规则引擎的循证医学知识库，是AI赋能教学的核心模块。它系统整合了疾病病因、诊断、治疗、预后等知识节点及其逻辑关系，构建了一个结构化的医学知识网络。该知识库整合了临床指南、诊疗评价和临床基础研究等核心循证证据，形成标准统一、体系完善的教学知识来源。系统的规则引擎能够依据患者症状生成初步诊断、推荐关键鉴别诊断路径，不仅提供符合国家标准的治疗方案作为参考，还能模拟临床医生的决策路径，将抽象的循证医学规则转化为可执行、可交互的医学教学逻辑。融合了全过程医疗知识图谱与规则引擎的循证医学AI知识库，可应用到临床系统架构教学场景，有助于学生直观理解循证知识内在关联，精准把握临床决策过程中的逻辑，提高反馈思维能力，快速掌握临床决策的“规则边界”，从而有效解决传统教学中知识点零散、逻辑难以串联的困境。这新型AI协同模式，能系统构建Kern教程，如PICO问题导向评估、循证教学的目标设定、特色化策略、过程实施、系统评估、教学质量反馈，更完善循证医学教学实践

[8]

。

2.3．DeepRare的“答案生成”到“思维播放”教学模式

2026年上海交通大学发布的DeepRare医学循证推理智能体，代表了AI赋能循证医学教学关键技术跨越。与通用大语言模型不同，DeepRare的核心突破在于其完整的“假设-验证-反思”推理闭环，能够呈现从患者表型到诊断结论的每一步推理路径。基于这一技术特征，我们有望将“推理显性化教学”（Reasoning Visualization Teaching）引入循证医学课堂，使具备可解释、可回溯推理能力的AI系统成为教学利器。这一转变的深远意义在于，它打破了传统思维“黑箱困境”，让学生能够直观跟进、深度参与、批判审视每一步临床推理，为从“授人以鱼”到“授人以渔”的教学转型提供了可能。此外，DeepRare结合循证医学，使其核心精髓变得更可见、可学、可批判，从而实现从“教知识”到“教思维”的根本转变。因此，我们认为DeepRare与循证医学的结合，有助于落实“知识-技能-态度”三维AI素养框架

[9]

，显著提升高校复合型医学人才的循证思维和推理能力。AI循证医学教学理论整合的框架见图1，常用AI辅助循证医学教学工具功能为表1。

Download: Download full-size image

Figure 1. 图1 AI辅助循证医学教学理论框架。

表1 常用AI辅助循证医学教学的工具和功能。

名称	功能	教学应用	优势	局限性
ChatGPT/GPT-4	自然语言对话、病例生成	证据检索辅助、PICO转化	交互自然、易获取	推理过程不透明、可能存在幻觉
知识图谱引擎	结构化知识推理	临床路径展示、规则教学	逻辑严谨、可溯源	交互性有限、更新滞后
DeepRare	推理显性化、假设-验证-反思闭环	思维训练、批判性评估	推理过程可见、可批判	专科覆盖有限、需专业部署
虚拟现实+GPT	沉浸式场景+对话	OSCE训练、模拟问诊	高仿真、体验感强	设备成本高、普及难度大

3．AI在循证医学教学中的应用

3.1．教学实践

Gupta等

[10]

人在系统综述中指出，生成式AI根据特定的知识点和学习需求，利用最新教学资源，快速生成大量高度逼真的临床病例，其技术的性能高度依赖于训练数据的质量。本研究在2025年9月—2026年3月期间，以本校2022-2023级临床医学专业学生为调查对象，在知情同意后采用计算机随机分组方案，将学生分为实验组（n=120）和对照组（n=120）。实验组接受AI辅助的循证医学教学，以本课题组“益生菌干预血脂研究Meta分析”

[11]

为内容。对照组接受传统讲授+案例讨论教学。调查内容如下：（1）AI推理与批判性分析；（2）AI生成证据的偏倚识别；（3）人机协同临床决策模拟。评价指标为：（1）证据检索效率（采用标准化PICO框架检索任务，由两名循证医学教师独立评分）；（2）批判性质疑频次（课堂录音转写后由研究助理编码统计）；（3）临床决策准确率（学生提交决策方案后由专家盲审评分）；（4）证据分级准确率（GRADE分级任务测试）；（5）循证实践总分（循证Fresno test量表，见表2）。结果显示，实验组在所有指标上均显著优于对照组（P<0.05）。本研究未记录学生AI使用频率与交互深度等过程数据，未分析不同基线水平或学习风格学生在AI辅助教学中的获益差异。未来应扩大样本量、延长干预时间、控制偏倚，并进行多中心亚组分析。

教育者尤为关切的是，如果AI所依赖的训练数据本身混杂了有设计缺陷或结论错误的医学研究，那么其生成的教学案例就可能披上“科学外衣”而传播伪科学，导致学生在模仿与学习中，不自觉地内化了错误的临床推理逻辑，从而使循证医学“以最佳证据为基础”的教学核心落空。此外，医学研究证据本身可能存在研究设计缺陷、样本选择偏差或数据分析错误、所生成的教学案例给学生传递错误信息等问题。为此，若要AI真正成为循证医学教学的可靠助手，就必须使其具备对自身知识盲区的敏感性，主动学习机制的优点。如面对存在偏倚、缺陷、错误的原始研究时，AI不应强行输出一个看似确定的答案，而应识别并标记出其中的不确定性。AI将这些标记样本交由技术专家进行规范化处理，这就相当于为教学AI配备了一位“监督员”。这种“AI识别存疑-专家反馈校准-模型迭代更新”的循环迭代训练，使AI不再是静态的知识库，而是一个能够持续从专家智慧中学习的动态教学辅助系统，其输出的可靠性、真实性也随之在每一次教学迭代中得到完善。如教学实践中，教师可以利用AI构建基于PICO框架的精准检索策略，学生则通过与AI的交互学习如何将临床问题转化为可检索的科学问题。

表2 AI辅助循证教学与传统教学效果对比结果。

指标	对照组（n=120）	实验组（n=120）	差异	t值	P值
证据检索效率（分）	61.3±8.7	85.5±6.3	+24.2	6.82	<0.001
批判性质疑频次（次/案例）	2.0±0.8	5.3±1.2	+3.2	7.45	<0.001
临床决策准确率（%）	67.4±10.2	84.2±7.1	+16.8	5.21	<0.001
证据分级准确率（%）	60.2±11.5	83.6±8.4	+23.4	6.58	<0.001
Fresno总分（满分120）	72.1±9.6	90.4±6.9	+18.3	5.96	<0.001

注：河南医药大学2022-2023级临床医学专业学生的AI辅助循证医学教学效果研究（n=120）。表中数据为均值±标准差。数据经正态性检验符合正态分布，采用独立样本t检验显示，实验组和对照组间差异均具有统计学意义（P<0.05）。

3.2．证据评估与批判性思维培养

人工智能的介入，使得循证医学教学的重点从“检索证据”转向“判断AI生成的证据”。有研究指出，若学生高度依赖AI生成的方案，可能导致其思维惰化或盲目信任AI，进而削弱其独立思考能力。为此，AI辅助教学必须将“批判性评估AI结果”作为新的核心技能，如DeepRare所代表的“推理显性化”范式下，“批判性评估AI结果”深入到AI的每一步推理逻辑，从而鼓励学生质疑“这个证据是否正确”，以及追问“AI为什么选该证据为最佳”。因此，AI辅助教学改革的方向应是持续利用AI构建高仿真训练环境，由教师引导学生主动质疑AI提供的案例与方案，并判断其逻辑与证据真伪，指导学生通过核查文献来源、交叉验证数据库等方法鉴别信息的真伪，从而帮助学生在今后个性化学习中确立决策主体地位。此外，DeepRare的“推理显性化”还发挥优势引入到“认知学徒制教学模式”和“自我调节学习理论及其教学实践”。这就更能模仿老师或专家思维，提升学生“自律”如“循证医学问题POCO”目标设定，循证证据收集、评估等实践过程监控，关键指标选取、统计结果解释及其临床意义等反思评价等，从而培养学生高阶思维能力和解决复杂问题的能力。

3.3．证据整合与应用

AI的辅助作用不仅体现在高效的信息查询和语言处理上，更直接表现为学生实践认知能力的提升。在教学方法上，AI赋能下的循证医学可构建独特的高仿真训练环境，例如敏捷循证设计（Agile Evidence-Informed Design，AVIDesign）模型

[12]

，这为信息查询与语言辅助提供了新的方法论框架。该模型紧密对应循证医学的实践步骤及课程设计系统的流程，使AI工具能够嵌入“从问题识别到解决方案”，并生成的完整教学链条。例如，在以I型超敏反应的临床常见疾病循证医学教学实践案例中，AI协同学生完成小组文献调研、讨论提纲拟定与课堂深度辩论等，结果显示学生对证据搜集、批判性评估到形成临床决策的全过程训练步骤的认识明显提高，有效提升了学生在真实循证医学场景中整合与运用证据的能力。此外，在AI技术特点上，它在循证医学教学过程中能够生成针对学生的学业质量报告、错题集、个人综合素质的分析报告，辅助学生及时了解自身学习状态同时，进行反思和判断。而对教师而言，AI能对其课堂教学质量评价进行分析与反馈，这为推动教学改革提供客观依据。

4．教学模式的重构与路径探索

4.1．传统课堂模式到AI辅助学生个性化学习

传统“一刀切”的教学模式，已难以适应医学生日益显著的个体差异。随着教学内容的深入，学生在认知水平和思维能力上的分化愈发明显，传统教学的局限性也因此凸显。AI融合的教学方案旨在满足学生的个体差异化需求。当前，尽管学生在先备知识、实践能力及对AI辅助教学的信任程度上存在显著差异，但普遍熟悉AI工具的基础操作。因此，AI辅助循证医学课程的设计必须从传统授课转向个性化学习，拟设计“理论基础+分层实践进阶”的教学路径，为不同起点的学生提供与其认知水平和实践能力相匹配的学习内容，支持和满足处于不同学生的学习需求。AI辅助课堂中，教师的核心在于构建“即时反馈、协作共学、动态评估”三位一体的教学机制，通过持续追踪学习进程并及时调整教学策略，推动教学模式从单向灌输走向精准适配。

4.2．“工具辅助”到“思维共振”的人机协同教学模式

AI技术正驱动循证医学教学从“人机分离”走向“人机协同”。最新报告显示，AI负责教学过程病例生成及证据收集、筛选等标准化任务，而教师则聚焦于逻辑解码与情感引导，这对“人-智协同教学”模型界定了一种新型分工方案。最近，以DeepRare为代表的新型AI系统，通过呈现“推理显性化路径”，将人机协同推向了“人机共同推理”的新层次。这使得学生能够随时与AI的每一步逻辑进行对话、辩驳，从而共同建构临床决策。该模式实现了人类导师的教学直觉与AI的效率优势的深度融合，为循证医学决策训练开辟了新路径，具有较好应用潜能。

4.3．互联网+循证教学为基础的产教融合

人工智能技术正驱动循证医学教学从封闭的传统课堂走向开放的“互联网+循证教学”新形态。山东大学齐鲁医院的实践表明，通过线上线下混合式教学探索，新增教学视频150分钟，线上选课人数突破1.4万人次，优质教学资源的辐射范围显著扩大。AI赋能教学的价值释放不止于课堂边界拓展，更在于其与宏观人才培养体系的深度融合。此外，借鉴高质量职业素质产学研协同人才培养模式，AI循证教学应紧紧与当地疾控中心、研究所、医疗企业、医院等实践基地深度整合，让学生在真实医学问题中完成“人机协同”的循证决策全流程训练，从而使AI辅助循证医学协同教学模式真正落地生根

[13]

。

4.4．国内外AI循证医学教育实践比较

2025年美国医学协会（AAMC）公布《医学教育中AI整合指南》，明确强调AI在教育中的伦理实施的“透明度、公平性、可问责”三原则

[14]

。英国的牛津大学循证医学中心（CEBM）要求医学生可以在导师指导下完成AI协同的系统评价，实践“做的过程可以学”与“研究过程可以学”的教学模式。我国广东医科大学建立AI医学院，引入20多个临床专病大模式，构建“边学边用”与“教评议题”的模式，并发布医学生AI素养能力清单的21项能力指标。我们认为，上述三国的AI素质培养的路径各异，美国侧重伦理能力框架设计，英国聚焦研究能力实践嵌入，我国强调产教融合与资源普惠设计。

5．关键问题

5.1．AI辅助教学技术可靠性

在AI辅助医学教学中，技术可靠性是确保教育质量与国家人才培养战略需求同频共振的基石。技术可靠性的核心在于确保传授给学生的知识是经过验证的，这是培养高素质医学人才的基石。首先，当前AI模型依赖的训练数据若存在循证医学某研究设计缺陷或样本偏差，将直接导致教学案例失真，不仅无法培养合格的医学人才，反而可能因错误知识的传播对未来的医疗安全构成潜在威胁。其次是“算法黑箱”风险（即AI的决策过程不透明，其内部机制难以被外界观察和理解，可能导致数据使用失控或生成与事实相悖的内容）与生成内容不可控的风险。因此，AI辅助循证医学教学过程，需要构建高质量、专用于循证医学教学的知识库，并建立数据质量的持续监控与更新机制。在可解释、可溯源的AI技术环境下，教育者能够清晰追溯AI的推理逻辑并进行有效纠偏，确保教学内容与国家立德树人的育人导向高度一致，强化教师在人才培养过程中的主导权。这不仅保障了技术底座的可靠性，更使得教学内容能够紧跟国家医学战略需求，实现前沿知识向课堂资源的快速转化，为拔尖创新人才的精准供给提供可持续的技术支撑。

5.2．AI辅助教学质量有效性

在AI教学工具快速迭代的背景下，其对学生核心素养的实际提升效果尚缺乏坚实的实证支撑，这是当前医学教育数字化转型中必须正视的关键短板。当前AI辅助教学的评估多依赖短期的满意度调查与学生自评，这类主观指标虽能反映课堂活跃度与接受度，却无法等同于学生核心素养的真实内化。学生主观感受上的“有帮助”，与其是否真正掌握了循证思维、能否独立解决复杂临床问题，存在本质差异。若停留于浅层评价，可能导致AI辅助教学改革陷入“技术繁荣、素养空心”的误区。为确保AI工具真正服务于人才培养目标，教学评估必须从即时的满意度调查转向更严谨的能力确证。应设计大样本、长周期、随机对照的实验研究，引入标准化的临床决策评估、循证实践考核等客观指标，系统观测AI干预对学生批判性思维、证据应用能力及临床推理质量的实际影响。只有通过高水平的循证研究验证其有效性，AI辅助教学工具才能从“辅助手段”真正升级为“育人利器”。这既是医学教育回应质量问责的必然要求，也是推动教学创新走向科学化、规范化的战略选择。

5.3．AI在医学教育中伦理、隐私与公平性问题

若训练数据本身隐含了对特定人群（如种族、性别、地域）的偏见，AI生成的教学案例便可能携带歧视性逻辑。这不仅向学生传递了错误的临床观念，更违背了医学教育培养具有公平意识、人文关怀的医务工作者的根本目标。此外，AI辅助教学应用可能加剧不同地区教育资源分布的不公平。例如，发达地区与顶尖院校凭借数据与技术优势率先受益于先进AI教学工具，而资源有限的地区则因基础设施与人才储备不足难以同步跟进。这种技术驱动下的教学资源分配不均，若不加干预，将直接导致不同院校医学生之间的教育质量差距被进一步拉大。这一趋势与国家推动医学教育均衡发展、促进优质资源共享的战略方向相悖。对此，教育主管部门需保持高度警觉，在鼓励技术创新的同时，应前瞻性地布局教育资源补偿机制与区域协同发展政策，确保AI技术成为缩小差距的桥梁，而非加剧分化的鸿沟。因此，应集中力量构建符合伦理规范的虚拟患者数据库，确保教学案例的多样性、代表性与科学性。通过标准化、高质量的虚拟临床资源，既保障了教学的规范性，也为不同地区的医学生提供了同等的优质实训机会，从源头上支撑医学教育的均衡发展。针对伦理与公平性问题，建议AI辅助循证医学教学实践应结合Beauchamp-Childress四原则（如自主、行善、不伤害、公正）和AI伦理原则（如透明性、可问责性），构建符合伦理规范的虚拟患者数据库，确保教学案例的多样性、代表性与科学性。

5.4．AI辅助教学的深层次育人价值问题

生成式AI提供答案的即时性与便捷性，若缺乏教学引导，容易使学生从“主动求索”滑向“被动接受”，国外学者同样警示，过度依赖AI生成结果可能削弱临床推理能力，侵蚀医学生与患者之间的人文联结

[15]

。这种依赖性若持续蔓延，将直接侵蚀医学生最核心的批判性思维与自主循证能力，导致技术越智能，学生的临床推理能力反而越薄弱。如何让学生在借助AI时不失独立思考，是教学设计必须回应的紧迫课题。归根结底，任何技术改革的成败都取决于“人”的因素，其核心在于能否实现深层次的育人价值。对于学生而言，主要风险在于“思维依赖”，既生成式人工智能提供答案的便捷性容易导致学生产生依赖，削弱其批判性思维和自主循证能力的培养。对于教师而言，从传统的单向知识传授者，转变为AI教学活动的设计者、引导者与监督者。这一转型对教师的数字素养、教学理念与课堂驾驭能力提出了全新要求。若缺乏系统性的培训支持与制度保障，教师极易陷入能力不足的困境，导致昂贵的AI工具因“不会用、不敢用”而被束之高阁，最终造成教育资源的巨大浪费。AI辅助循证医学教育过程，关键在于坚持“育人为本”：一方面，要通过教学设计引导学生与技术保持恰当的批判性距离，强化其独立思考能力；另一方面，要为教师提供分层分类的精准培训与长效支持机制，帮助其顺利实现角色转型。

6．未来展望

6.1．技术优化与专科工具开发

当前通用生成式AI工具在循证医学教学中存在一定局限性，如训练数据往往部分缺乏循证医学证据的深度和针对性。为此，继续深入开发基于高质量循证医学证据的专用AI系统，是提升该教学专业性和准确性的关键路径。未来应致力于建立专科AI教学工具与动态知识更新机制，推动传统知识库向智能化的“教学知识库2.0”升级，将循证医学证据体系转化为机器可理解、可计算的先验知识框架。

6.2．课程体系重构与评估标准建立

在AI辅助课程体系，可增设“AI循证工具实践课”等新课程，推动医教结合的教育改革，培养学生运用AI工具解决临床实际问题的能力。最新研究提出了AI能力的层级框架，包括认知能力（理解AI原理）、操作能力（恰当使用AI工具）和元AI能力（为临床情境选择最佳AI模型），为本课程体系的构建提供了理论参考

[16]

。在评估体系构建方面，可借鉴PBL（问题导向学习）理念，建立科学的AI教学工具评估体系，构建“学评融合”的新型评价机制。这种机制能够为不同教学场景提供个性化的评价方案，推动教学评价从主观满意度调查向客观量化评价转变，从而实现教学质量的持续改进。

6.3．教师发展与资源均衡

AI协同的医学教育改革面临多学科交叉人才匮乏、配套资源及服务体系滞后的困境，而产学研协同的人才培养模式为此提供了有效的破解路径。针对这一问题，需要以教师能力建设为突破口，以资源均衡配置为支撑，以循证研究深化为保障，构建可持续发展的教学生态。教师是推动教学变革的第一资源。面对“人工智能+医学教育”的转型需求，应系统开展专项培训，帮助教师掌握运用AI工具开展循证教学的核心技能，顺利实现从知识传授者向教学活动设计者、引导者的角色转型。同时，建立分层分类的师资认证机制，通过标准化能力认证体系，打造一支既懂医学又懂AI、既精教学又善创新的复合型师资队伍，从源头夯实教学改革的人才基础。未来研究应聚焦以下方向：（1）在师资培养的基础上，教育主管部门应发挥政策引导作用，通过专项扶持推动优质AI教学资源向欠发达地区有序流动。鼓励发达地区与资源薄弱院校建立对口合作机制，实现虚拟仿真案例库、循证医学数据库等优质资源的共建共享。（2）在资源共享的基础上，探索线上线下混合教学新形态：如由输出校专家担任主讲教师负责核心内容讲授，由接收校教师担任辅导教师负责线下互动与实训指导。这一“双师模式”既能发挥名师引领作用，又能确保教学落地实效，为不同地区学生提供同等高质量的循证医学教育。教师发展与资源均衡的成效，需要通过严谨的循证研究来验证。当前AI辅助教学的评估多依赖短期满意度调查，对学生核心素养提升效果的实证支撑尚不充分。下一步研究应设计大样本、长周期的随机对照实验，引入标准化临床决策评估、循证实践考核等客观指标，系统观测AI干预对学生批判性思维、证据应用能力及临床推理质量的实际影响。研究成果可反哺教学优化与资源配置，为AI辅助循证医学教学模式的可持续发展提供依据。因此，新时代高素质医学人才的培养要求，不仅统筹推进医学教育智能化升级、促进多方协作以外，须形成“教师-学生-AI”三位一体的新型教育关系，才能显著提升高校医学生的循证思维和推理能力。

综上，本研究结合国内外最新文献，聚焦AI技术与循证医学教学深度融合的前沿探索过程，系统阐述AI在循证医学教学中三螺旋模型、Kern六步课程框架、认知学徒制理论与三维AI素养框架整合为统一的AI循证医学教学分析框架。本研究认为以DeepRare的“假设-验证-反思”推理闭环为技术基础，将传统教学中不可见的临床思维过程转化为可观察、可模仿、可批判的教学内容，实现了从“教知识”到“教思维”的范式转换。今后，AI辅助循证医学实践过程，开发低成本、易接触使用的AI教学工具，提升智能教学方案的可推广性。本研究皆为适合我国医学发展更深入智能化循证医学教育转型实践提供理论依据。

References

[1]	Looi K M. Can AI teach medicine? [J]. BMJ (Clinical research ed.), 2025, 389r822. https://doi.org/10.1136/BMJ.R822
[2]	Zhang Q, Huang Z, Huang Y, et al.Generative AI in medical education: feasibility and educational value of LLM-generated clinical cases with MCQs. [J]. BMC medical education, 2025, 25(1): 1502. https://doi.org/10.1186/S12909-025-08085-8
[3]	Mehraeen E, Khademzadeh S, Habibi P, et al. New technologies in medical education: An umbrella review on recent evidence [J]. J Educ Health Promot, 2026, 15(1): 26. https://doi.org/10.4103/jehp.jehp_1462_24
[4]	Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019; 25(1): 44-56. https://doi.org/10.1038/s41591-018-0300-7
[5]	Gurbuz S, Karslioglu B, Keskin A, Igde N, Ayaz MB, Imren Y. Comparative Efficacy of ChatGPT and Gemini in Addressing Patient Queries on Gonarthrosis and Total Knee Arthroplasty: A Randomized Controlled Trial. J Knee Surg. 2026; 39(3): 123-126. https://doi.org/10.1055/a-2693-0756
[6]	Thompson RAM, Shah YB, Aguirre F, Stewart C, Lallas CD, Shah MS. Artificial Intelligence Use in Medical Education: Best Practices and Future Directions. Curr Urol Rep. 2025; 26(1): 45. Published 2025 May 29. https://doi.org/10.1007/s11934-025-01277-1
[7]	苏友利, 苗长城, 程轶. 智能技术的教育适配性研究——基于DeepSeek R1的临床思维培养路径重构 [J]. 延边大学医学学报, 2025, 48(9): 128-131. https://doi.org/10.16068/j.1000-1824.2025.09.037
[8]	Ahsan Z. Integrating artificial intelligence into medical education: a narrative systematic review of current applications, challenges, and future directions. BMC Med Educ. 2025; 25(1): 1187. Published 2025 Aug 23. https://doi.org/10.1186/s12909-025-07744-0
[9]	张思玮. 人工智能时代, 如何培育复合型医学人才? [N]. 中国科学报, 2025-09-03(003). https://doi.org/10.28514/n.cnki.nkxsb.2025.003380
[10]	Gupta DK, Chaudhuri A. Artificial Intelligence and Digital Technologies in Medical Education: A Systematic Review of Effectiveness, Equity, and Future Directions [J]. J Mar Med Soc, 2026. https://doi.org/10.4103/jmms.jmms_144_25
[11]	朱瑜, 朱绍辉, 郑志, 等. 益生菌干预方案改善肥胖人群血脂水平的Meta分析 [J]. 护理研究. 2024, 38 (01): 165-170.
[12]	Daroowalla F, Ratliff M, Migriauli I, Hirumi A. Agile EVidence-Informed Design (AVIDesign): Transdisciplinary Curriculum Design for Evidence-Informed Health Professions Education. Med Sci Educ. 2025;35(5):2291-2300. Published 2025 Jun 5. https://doi.org/10.1007/s40670-025-02418-4
[13]	郑志, 陈璇, 李守良, 等. 高质量MPH医学人才培养产学研协同创新教育模式探索与实践 [J]. 高校医学教学研究(电子版), 2025, 15(02): 17-24.
[14]	Itani A, Gronseth SL, Musaad S, Nguyen T, Mirabile Y, Beech BM. Ethical considerations for teaching with artificial intelligence: a scoping review in medical education settings. Int J Educ Technol High Educ. 2025;22(1):68. https://doi.org/10.1186/s41239-025-00563-9
[15]	Saroha S. Artificial Intelligence in Medical Education: Promise, Pitfalls, and Practical Pathways. Adv Med Educ Pract. 2025;16:1039-1046. Published 2025 Jun 14. https://doi.org/10.2147/AMEP.S523255
[16]	Tsuei S. AI Competency: Current State and Challenges. JMIR Med Educ. 2026; 12: e86686. Published 2026 Mar 3. https://doi.org/10.2196/86686

Cite This Article

Plain Text BibTeX RIS

APA Style

Qian, D., Ke, W., Zhi, Z. (2026). Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching. Science Research, 14(4), 253-260. https://doi.org/10.11648/j.sr.20261404.26

Copy | Download

ACS Style

Qian, D.; Ke, W.; Zhi, Z. Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching. Sci. Res. 2026, 14(4), 253-260. doi: 10.11648/j.sr.20261404.26

Copy | Download

AMA Style

Qian D, Ke W, Zhi Z. Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching. Sci Res. 2026;14(4):253-260. doi: 10.11648/j.sr.20261404.26

Copy | Download

@article{10.11648/j.sr.20261404.26,
  author = {Dai Qian and Wei Ke and Zheng Zhi},
  title = {Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching},
  journal = {Science Research},
  volume = {14},
  number = {4},
  pages = {253-260},
  doi = {10.11648/j.sr.20261404.26},
  url = {https://doi.org/10.11648/j.sr.20261404.26},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.sr.20261404.26},
  abstract = {Modern generative artificial intelligence technologies are fundamentally reshaping the pedagogical paradigm of evidence-based medicine (EBM). This article systematically articulates the mechanisms by which AI augments EBM instruction—encompassing large language models, knowledge graphs, and rule engines—with a specific focus on the "reasoning visualization" pedagogical approach introduced by emerging medical evidence-based AI agents such as DeepRare. In contrast to the functional orientation of general-purpose large language models, DeepRare operationalizes a complete "hypothesis-verification-reflection" reasoning loop, thereby transforming the implicit clinical reasoning processes of conventional instruction into observable, replicable, and critically assessable educational content. Building upon this foundation, the article dissects the concrete pathways of AI integration in EBM teaching across four interdependent dimensions: evidence acquisition and screening, evidence appraisal and the cultivation of critical thinking, evidence synthesis and clinical translation, and teaching evaluation with feedback mechanisms. The inquiry further explores the reconceptualization of teaching models, encompassing competency-adaptive personalized learning trajectories, human-computer collaborative decision-making exercises that promote shared clinical reasoning between learners and intelligent systems, and the "Internet plus evidence-based teaching" paradigm that extends educational impact beyond traditional classroom settings. In confronting practical challenges—including technological dependability, the paucity of empirical validation regarding educational outcomes, ethical and equity considerations, and the digital competency gaps among both instructors and learners—this article advances an integrated four-pronged future agenda comprising technological refinement, curricular redesign, faculty professional development, and deepened empirical research. The ultimate objective is to furnish an operational foundation for the intelligent transformation of EBM education that is congruent with the evolving demands of medical development and talent cultivation in China.},
 year = {2026}
}

Copy | Download

TY  - JOUR
T1  - Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching
AU  - Dai Qian
AU  - Wei Ke
AU  - Zheng Zhi
Y1  - 2026/08/26
PY  - 2026
N1  - https://doi.org/10.11648/j.sr.20261404.26
DO  - 10.11648/j.sr.20261404.26
T2  - Science Research
JF  - Science Research
JO  - Science Research
SP  - 253
EP  - 260
PB  - Science Publishing Group
SN  - 2329-0927
UR  - https://doi.org/10.11648/j.sr.20261404.26
AB  - Modern generative artificial intelligence technologies are fundamentally reshaping the pedagogical paradigm of evidence-based medicine (EBM). This article systematically articulates the mechanisms by which AI augments EBM instruction—encompassing large language models, knowledge graphs, and rule engines—with a specific focus on the "reasoning visualization" pedagogical approach introduced by emerging medical evidence-based AI agents such as DeepRare. In contrast to the functional orientation of general-purpose large language models, DeepRare operationalizes a complete "hypothesis-verification-reflection" reasoning loop, thereby transforming the implicit clinical reasoning processes of conventional instruction into observable, replicable, and critically assessable educational content. Building upon this foundation, the article dissects the concrete pathways of AI integration in EBM teaching across four interdependent dimensions: evidence acquisition and screening, evidence appraisal and the cultivation of critical thinking, evidence synthesis and clinical translation, and teaching evaluation with feedback mechanisms. The inquiry further explores the reconceptualization of teaching models, encompassing competency-adaptive personalized learning trajectories, human-computer collaborative decision-making exercises that promote shared clinical reasoning between learners and intelligent systems, and the "Internet plus evidence-based teaching" paradigm that extends educational impact beyond traditional classroom settings. In confronting practical challenges—including technological dependability, the paucity of empirical validation regarding educational outcomes, ethical and equity considerations, and the digital competency gaps among both instructors and learners—this article advances an integrated four-pronged future agenda comprising technological refinement, curricular redesign, faculty professional development, and deepened empirical research. The ultimate objective is to furnish an operational foundation for the intelligent transformation of EBM education that is congruent with the evolving demands of medical development and talent cultivation in China.
VL  - 14
IS  - 4
ER  -

Copy | Download

Author Information

Dai Qian

School of Public Health, Henan Medical University, Xinxiang, China;Xinxiang Key Laboratory of Applied Microbial Technology, Xinxiang, China
Wei Ke

School of Public Health, Henan Medical University, Xinxiang, China;Xinxiang Key Laboratory of Applied Microbial Technology, Xinxiang, China
Zheng Zhi

School of Public Health, Henan Medical University, Xinxiang, China;Xinxiang Key Laboratory of Applied Microbial Technology, Xinxiang, China;Xinxiang Key Laboratory of Treatment and Medical Transformation of Obesity and Type 2 Diabetes, the First Affiliated Hospital of Henan Medical University, Weihui, China

Contact Email

Download PDF

Submit an Article

Figure 1

Figure 1. 图 1 AI辅助循证医学教学理论框架。

Table 1

表1 常用AI辅助循证医学教学的工具和功能。
Table 2

表2 AI辅助循证教学与传统教学效果对比结果。

Plain Text BibTeX RIS

APA Style

Qian, D., Ke, W., Zhi, Z. (2026). Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching. Science Research, 14(4), 253-260. https://doi.org/10.11648/j.sr.20261404.26

Copy | Download

ACS Style

Qian, D.; Ke, W.; Zhi, Z. Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching. Sci. Res. 2026, 14(4), 253-260. doi: 10.11648/j.sr.20261404.26

Copy | Download

AMA Style

Qian D, Ke W, Zhi Z. Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching. Sci Res. 2026;14(4):253-260. doi: 10.11648/j.sr.20261404.26

Copy | Download

@article{10.11648/j.sr.20261404.26,
  author = {Dai Qian and Wei Ke and Zheng Zhi},
  title = {Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching},
  journal = {Science Research},
  volume = {14},
  number = {4},
  pages = {253-260},
  doi = {10.11648/j.sr.20261404.26},
  url = {https://doi.org/10.11648/j.sr.20261404.26},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.sr.20261404.26},
  abstract = {Modern generative artificial intelligence technologies are fundamentally reshaping the pedagogical paradigm of evidence-based medicine (EBM). This article systematically articulates the mechanisms by which AI augments EBM instruction—encompassing large language models, knowledge graphs, and rule engines—with a specific focus on the "reasoning visualization" pedagogical approach introduced by emerging medical evidence-based AI agents such as DeepRare. In contrast to the functional orientation of general-purpose large language models, DeepRare operationalizes a complete "hypothesis-verification-reflection" reasoning loop, thereby transforming the implicit clinical reasoning processes of conventional instruction into observable, replicable, and critically assessable educational content. Building upon this foundation, the article dissects the concrete pathways of AI integration in EBM teaching across four interdependent dimensions: evidence acquisition and screening, evidence appraisal and the cultivation of critical thinking, evidence synthesis and clinical translation, and teaching evaluation with feedback mechanisms. The inquiry further explores the reconceptualization of teaching models, encompassing competency-adaptive personalized learning trajectories, human-computer collaborative decision-making exercises that promote shared clinical reasoning between learners and intelligent systems, and the "Internet plus evidence-based teaching" paradigm that extends educational impact beyond traditional classroom settings. In confronting practical challenges—including technological dependability, the paucity of empirical validation regarding educational outcomes, ethical and equity considerations, and the digital competency gaps among both instructors and learners—this article advances an integrated four-pronged future agenda comprising technological refinement, curricular redesign, faculty professional development, and deepened empirical research. The ultimate objective is to furnish an operational foundation for the intelligent transformation of EBM education that is congruent with the evolving demands of medical development and talent cultivation in China.},
 year = {2026}
}

Copy | Download

TY  - JOUR
T1  - Exploration of the Application Mode of Artificial Intelligence Assisted Evidence Based Medicine Collaborative Teaching
AU  - Dai Qian
AU  - Wei Ke
AU  - Zheng Zhi
Y1  - 2026/08/26
PY  - 2026
N1  - https://doi.org/10.11648/j.sr.20261404.26
DO  - 10.11648/j.sr.20261404.26
T2  - Science Research
JF  - Science Research
JO  - Science Research
SP  - 253
EP  - 260
PB  - Science Publishing Group
SN  - 2329-0927
UR  - https://doi.org/10.11648/j.sr.20261404.26
AB  - Modern generative artificial intelligence technologies are fundamentally reshaping the pedagogical paradigm of evidence-based medicine (EBM). This article systematically articulates the mechanisms by which AI augments EBM instruction—encompassing large language models, knowledge graphs, and rule engines—with a specific focus on the "reasoning visualization" pedagogical approach introduced by emerging medical evidence-based AI agents such as DeepRare. In contrast to the functional orientation of general-purpose large language models, DeepRare operationalizes a complete "hypothesis-verification-reflection" reasoning loop, thereby transforming the implicit clinical reasoning processes of conventional instruction into observable, replicable, and critically assessable educational content. Building upon this foundation, the article dissects the concrete pathways of AI integration in EBM teaching across four interdependent dimensions: evidence acquisition and screening, evidence appraisal and the cultivation of critical thinking, evidence synthesis and clinical translation, and teaching evaluation with feedback mechanisms. The inquiry further explores the reconceptualization of teaching models, encompassing competency-adaptive personalized learning trajectories, human-computer collaborative decision-making exercises that promote shared clinical reasoning between learners and intelligent systems, and the "Internet plus evidence-based teaching" paradigm that extends educational impact beyond traditional classroom settings. In confronting practical challenges—including technological dependability, the paucity of empirical validation regarding educational outcomes, ethical and equity considerations, and the digital competency gaps among both instructors and learners—this article advances an integrated four-pronged future agenda comprising technological refinement, curricular redesign, faculty professional development, and deepened empirical research. The ultimate objective is to furnish an operational foundation for the intelligent transformation of EBM education that is congruent with the evolving demands of medical development and talent cultivation in China.
VL  - 14
IS  - 4
ER  -

Copy | Download