Li Xin, Liu Haoyang, Zhao Xu, Zhang Guigang, Xing Chunxiao
1Department of Rehabilitation, Beijing Tsinghua Changgung Hospital Medical Center, Tsinghua University, Beijing, China.
2Beijing University of Posts and Telecommunications, Beijing, China.
Health Inf Sci Syst. 2020 Feb 27;8(1):12. doi: 10.1007/s13755-020-0102-4. eCollection 2020 Dec.
In this study, a medical knowledge graph is constructed from the electronic medical record text of knee osteoarthritis patients to support intelligent medical applications such as knowledge retrieval and decision support, and to promote the sharing of medical resources. After constructing the domain ontology of knee osteoarthritis and manually labeling, we trained a machine learning model to automatically perform entity recognition and entity relation extraction, and then used a graph database to construct the knowledge graph of knee osteoarthritis. The experiment proves that the knowledge graph is comprehensive and reliable, and the knowledge graph construction method proposed in this study is effective.
在本研究中，从膝骨关节炎患者的电子病历文本构建医学知识图谱，以支持诸如知识检索和决策支持等智能医疗应用，并促进医疗资源的共享。在构建膝骨关节炎的领域本体并进行人工标注后，我们训练了一个机器学习模型来自动执行实体识别和实体关系提取，然后使用图数据库构建膝骨关节炎的知识图谱。实验证明，该知识图谱是全面且可靠的，本研究提出的知识图谱构建方法是有效的。
