医学知识图谱构建研究进展

Advances in studies on construction of medical knowledge graphs

针对医学数据专业性强、结构复杂等特点,解析了构建医学知识图谱的关键技术,介绍了利用机器学习和深度学习的方法识别医学命名实体、实体链接和抽取语义关系,以及医学知识图谱在医院智能导诊、疾病筛查和预测、辅助临床诊断、医疗保险风险预测和医学知识科普方面的应用。结合当前我国医学知识图谱构建在数据和技术层面临的问题和挑战,提出了相应的对策和建议。

The key technology for the construction of medical knowledge graphs was analyzed in accordance with the characteristics of medical data such as strong specialization and complex structure with a description of the identification of entity that is named according to medical sciences,links between entities,extraction of semantic relationships by making use of machine learning and deep learning respectively,and the application of medical knowledge graphs in intelligence-guided diagnosis,screening and prediction of diseases,auxiliary clinical diagnosis of diseases,risk prediction of medical insurance,and spreading of popular medical knowledge.Countermeasures and suggestions were put forward for the problems and challenges that are faced by the data and technology in the construction of medical knowledge graphs in our country.