静脉血栓栓塞症智能知识库的构建及在医学教育中的应用评价

佘康; 郝雨迪; 王亚青; 李俊梅; 马颖林; 母光妍; 李珂; 张晓琳; 齐建光; 齐心

doi:10.3760/cma.j.cn116021-20260128-02296

您当前的位置：

首页 >

文章列表页 >

静脉血栓栓塞症智能知识库的构建及在医学教育中的应用评价

数智医学教育 | 更新时间：2026-06-17

- 静脉血栓栓塞症智能知识库的构建及在医学教育中的应用评价
- Construction of an intelligent knowledge base for venous thromboembolism and evaluation of its application in medical education
- 中华医学教育探索杂志 2026年25卷第6期页码：664-673
- 作者机构：
  
  1.北京大学第一医院介入血管外科，北京 100034
  2.国家癌症中心/国家肿瘤临床医学研究中心/中国医学科学院北京协和医学院肿瘤医院药剂科，北京 100021
  3.北京大学第一医院药剂科，北京 100034
  4.北京大学第一医院教育处，北京 100034
  5.北京大学第一医院烧伤整形科，北京 100034
- 作者简介：
  
  齐心，Email：randyq@126.com
- 基金信息：
  
  北京大学教学新思路3.0立项资助(2026XSL10);北京大学第一医院院级教学课题(2025YJ09)
- DOI：10.3760/cma.j.cn116021-20260128-02296
  中图分类号： R-4
- 收稿：2026-01-28，
  
  纸质出版：2026-06-20
- 稿件说明：
移动端阅览
佘康, 郝雨迪, 王亚青, 等. 静脉血栓栓塞症智能知识库的构建及在医学教育中的应用评价[J]. 中华医学教育探索杂志, 2026, 25(6): 664-673.

She Kang, Hao Yudi, Wang Yaqing, et al. Construction of an intelligent knowledge base for venous thromboembolism and evaluation of its application in medical education[J]. Chinese Journal of Medical Education Research, 2026, 25(6): 664-673.
佘康, 郝雨迪, 王亚青, 等. 静脉血栓栓塞症智能知识库的构建及在医学教育中的应用评价[J]. 中华医学教育探索杂志, 2026, 25(6): 664-673. DOI： 10.3760/cma.j.cn116021-20260128-02296.

She Kang, Hao Yudi, Wang Yaqing, et al. Construction of an intelligent knowledge base for venous thromboembolism and evaluation of its application in medical education[J]. Chinese Journal of Medical Education Research, 2026, 25(6): 664-673. DOI： 10.3760/cma.j.cn116021-20260128-02296.

摘要

目的

构建基于大语言模型的检索增强生成技术的静脉血栓栓塞症（venous thromboembolism，VTE）专业知识库，并评价其在医学教育中的应用效果。

方法

采用“循证证据—专家经验—真实病例”三层语料，在腾讯ima平台搭建VTE知识库，并实施证据分级、版本控制和人机协同审核。采用混合方法研究：学习者路径通过横断面调查评价医学生及非专科医护人员（

=131）在临床决策支持、知识问答和资源推荐中的使用体验；教师路径采用单盲随机对照设计，评价血管外科教师（

=19）在教案设计、病例讨论和辅助命题中的应用效果。统计分析在R 4.5.0软件中完成；计数资料采用频数和构成比描述，Likert量表评分采用中位数和四分位间距描述，组间比较采用Kruskal-Wallis

检验、Friedman秩和检验及Wilcoxon符号秩检验。

结果

学习者路径显示，知识库在标准化任务中应用效果较好，各维度满意度评分的中位数均为4.00分[四分位间距（interquartile range，IQR）为4.00~5.00

]

，但循证可信度评分相对较低。教师路径显示，其应用效果具有任务依赖性：在“教案设计”等收敛性任务中，深度使用知识库的适用性略高于入门使用；在“辅助命题”等创造性任务中，深度使用知识库的适用性低于入门使用（

0.05）。

结论

基于检索增强生成（retrieval-augmented generation，RAG）技术构建的VTE专业知识库在医学教育中具有一定应用价值，但其应用边界尚不明确。未来应通过构建涵盖教学法、亚专科及科研数据的“知识库矩阵”，进一步提升其在真实医学教育场景中的应用效果。

Abstract

Objective

To construct a venous thromboembolism (VTE) knowledge base based on retrieval-augmented generation powered by a large language model

and evaluate its effectiveness in medical education.

Methods

A three-layer corpus consisting of evidence-based information

expert experience

and real-world cases was used to build the VTE knowledge base on the Tencent IMA platform. Evidence grading

version control

and human-AI collaborative review were performed. A mixed-methods design was used. For the learner pathway

a cross-sectional survey was used to evaluate the experience of medical students and non-specialist healthcare workers (

=131) in clinical decision support

knowledge question answering

and resource recommendation. For the teacher pathway

a single-blind randomized controlled design was used to evaluate the effectiveness of application among vascular surgery teachers (

=19) in lesson planning

case discussion

and assisted exam item generation. Statistical analyses were performed using R 4.5.0. Categorical data were described using frequencies and proportions; Likert scale scores were described using medians and interquartile ranges. Inter-group comparisons were performed using the Kruskal-Wallis

test

Friedman test

and Wilcoxon signed-rank test.

Results

The learner pathway showed that the knowledge base performed well in standardized tasks

with median satisfaction scores of 4.00 (interquartile range: 4.00-5.00) across dimensions

although the evidence-based credibility score was relatively low. The teacher pathway showed that the application effectiveness was task-dependent. For convergent tasks such as lesson planning

intensive use of the knowledge base had slightly higher applicability than basic use. For c

reative tasks such as assisted exam item generation

intensive use had lower applicability than basic use (

0.05).

Conclusions

The VTE knowledge base constructed based on retrieval-augmented generation has certain application value in medical education

but its application boundaries remain unclear. Future efforts should focus on constructing a knowledge base matrix that integrates pedagogy

subspecialty knowledge

and research data to further improve its application effectiveness in real medical education settings.

关键词

Keywords

references

Wendelboe A , Weitz JI . Global health burden of venous thromboembolism [J ] . Arterioscler Thromb Vasc Biol , 2024 , 44 ( 5 ): 1007 - 1011 . DOI: 10.1161/atvbaha.124.320151 http://dx.doi.org/10.1161/atvbaha.124.320151 .

Henke PK , Kahn SR , Pannucci CJ , et al . Call to action to prevent venous thromboembolism in hospitalized patients: a policy statement from the American heart association [J ] . Circulation , 2020 , 141 ( 24 ): e914 - e931 . DOI: 10.1161/cir.0000000000000769 http://dx.doi.org/10.1161/cir.0000000000000769 .

Zhang S , Liu L , Ding S , et al . Mediator effects of cognitive load on the relationship between task complexity and guideline adherence among clinical nurses: a cross-sectional survey of nurses in China [J ] . J Res Nurs , 2025 , 30 ( 5/6 ): 479 - 490 . DOI: 10.1177/17449871251329197 http://dx.doi.org/10.1177/17449871251329197 .

田蕾，孙宝志 . 第四代医学教育考核评价：精准教育评价 [J ] . 中华医学教育探索杂志， 2025 ， 24 （ 11 ）： 1441 - 1446 . DOI: 10.3760/cma.j.cn116021-20250401-02127 http://dx.doi.org/10.3760/cma.j.cn116021-20250401-02127 .

徐赫，焦洋 . 人工智能在临床推理教学新模式中的研究及展望 [J ] . 中华医学教育探索杂志， 2025 ， 24 （ 10 ）： 1372 - 1378 . DOI: 10.3760/cma.j.cn116021-20250310-02066 http://dx.doi.org/10.3760/cma.j.cn116021-20250310-02066 .

Karajizadeh M , Hassanipour S , Sharifian R , et al . The effect of information technology intervention on using appropriate VTE prophylaxis in non-surgical patients: a systematic review and meta-analysis [J ] . Digit Health , 2022 , 8 : 205520762211188 . DOI: 10.1177/20552076221118828 http://dx.doi.org/10.1177/20552076221118828 .

Rao A , Pang M , Kim J , et al . Assessing the utility of ChatGPT throughout the entire clinical workflow: development and usability study [J ] . J Med Internet Res , 2023 , 25 : e48659 . DOI: 10.2196/48659 http://dx.doi.org/10.2196/48659 .

Janumpally R , Nanua S , Ngo A , et al . Generative artificial intelligence in graduate medical education [J ] . Front Med , 2025 , 11 : 1525604 . DOI: 10.3389/fmed.2024.1525604 http://dx.doi.org/10.3389/fmed.2024.1525604 .

Neha F , Bhati D , Shukla DK . Retrieval-augmented generation (RAG) in healthcare: a comprehensive review [J ] . AI , 2025 , 6 ( 9 ): 226 . DOI: 10.3390/ai6090226 http://dx.doi.org/10.3390/ai6090226 .

Liu S , McCoy AB , Wright A . Improving large language model applications in biomedicine with retrieval-augmented generation: a systematic review, meta-analysis, and clinical development guidelines [J ] . J Am Med Inform Assoc , 2025 , 32 ( 4 ): 605 - 615 . DOI: 10.1093/jamia/ocaf008 http://dx.doi.org/10.1093/jamia/ocaf008 .

唐江桥，戴珑珑 . 人工智能辅助教学的边界：高校教师的专业判断与价值坚守 [J ] . 黑龙江高教研究， 2026 ， 44 （ 1 ）： 13 - 18 . DOI: 10.19903/j.cnki.CN23-1074/G.2026.01.003 http://dx.doi.org/10.19903/j.cnki.CN23-1074/G.2026.01.003 .

关旭，刘骞，王锡山 . 人工智能在肿瘤外科医学专科教育中的现状与应用研究 [J ] . 中华医学教育探索杂志， 2024 ， 23 （ 10 ）： 1337 - 1342 . DOI: 10.3760/cma.j.cn116021-20240511-01869 http://dx.doi.org/10.3760/cma.j.cn116021-20240511-01869 .

Gharaibeh L , Albsoul-Younes A , Younes N . Evaluation of VTE prophylaxis in an educational hospital: comparison between the institutional guideline (caprini 2006) and the ACCP guideline (ninth edition) [J ] . Clin Appl Thromb Hemost , 2016 , 22 ( 7 ): 627 - 632 . DOI: 10.1177/1076029615575344 http://dx.doi.org/10.1177/1076029615575344 .

徐锐，郑兰斌，张燕蕾，等 . 生成式AI在医学教育中的应用：影响机制与风险应对 [J ] . 中国卫生事业管理， 2025 ， 42 （ 12 ）： 1398 - 1401， 1411 . DOI: 10.3969/j.issn.1004-4663.2025.12.014 http://dx.doi.org/10.3969/j.issn.1004-4663.2025.12.014 .

von Gerich H , Chomutare T , Kytö V , et al . A rapid review on the application of common data models in healthcare: recommendations for data governance and federated learning in artificial intelligence development [J ] . Digit Health , 2025 , 11 : 20552076251395536 . DOI: 10.1177/20552076251395536 http://dx.doi.org/10.1177/20552076251395536 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

大语言模型在公共卫生与预防医学专业院校教育中的应用与展望

DeepSeek 辅助命题用于前列腺癌教学测评的可行性研究

中国精英教学医院联盟医学教师数智胜任力框架共识

医学教学对象人工智能使用现状及伦理认知调查研究

中美临床药师核心胜任力框架比较