《生命科学》 2025, 37(12): 1505-1516

生物医学知识图谱辅助循证医学决策

井梦甜¹ , 雷　蕾¹ , 范海巍^1,2,*

¹上海科技大学，上海 201210 ²上海临床研究中心，上海 201210

摘　要：

循证医学决策面临信息过载与知识碎片化挑战。生物医学知识图谱作为新兴技术，展现了解决这些问题的潜力。本文系统性地分析知识图谱在循证医学中的应用现状与技术路径，揭示其如何改变临床证据的获取、整合与应用模式。知识图谱通过语义网络结构实现医学知识系统化表示与智能推理，为个性化治疗、罕见病诊断等提供决策支持。尽管面临证据质量评估、动态更新与可解释性等挑战，但知识图谱正推动循证医学从“基于文献”向“基于知识”转型，在证据整合与人机协同决策方面展现广阔前景。

通讯作者：范海巍 , Email:fanhw@shanghaitech.edu.cn

Biomedical knowledge graphs supporting evidence-based medicine decision-making

JING Meng-Tian¹ , LEI Lei¹ , FAN Hai-Wei^1,2,*

¹ShanghaiTech University, Shanghai 201210, China ²Shanghai Clinical Research and Trial Center, Shanghai 201210, China

Abstract:

Evidence-based medicine decision-making faces challenges of information overload and knowledge fragmentation. As an emerging technology, biomedical knowledge graphs show the potential to address these issues. This paper systematically analyzes the current application status and technical paths of knowledge graphs in evidence-based medicine, and reveals how they transform the acquisition, integration, and application models of clinical evidence. Knowledge graphs achieve systematic representation and intelligent reasoning of medical knowledge through semantic network structures, providing decision-making support for personalized treatment, rare disease diagnosis, etc. Although they face challenges such as evidence quality assessment, dynamic updating, and interpretability, knowledge graphs are driving the transformation of evidence-based medicine from "literaturebased" to "knowledge-based", and show broad prospects in evidence integration and human-machine collaborative
decision-making.

Communication Author：FAN Hai-Wei , Email:fanhw@shanghaitech.edu.cn