医学检验知识的图谱构建与应用  

The Construction and Application of Medical Examination Knowledge Graph Based on Laboratory Information System

作　　者：金雅妮 王华珍[1] 王硕 刘骏龙 JIN Yani;WANG Huazhen;WANG Shuo;LIU Junlong(School of Computer Science and Technology,Huaqiao University,Xiamen,China,361021)

机构地区：[1]华侨大学计算机科学与技术学院

出　　处：《福建电脑》2020年第2期1-7,共7页Journal of Fujian Computer

基　　金：国家自然科学基金项目(No.61673186);福建省自然科学基金项目(No.2012J01274);泉州市科技计划项目(No.2017G019)资助

摘　　要：为填补未有基于检验科信息系统(Laboratory information system,LIS)真实检验数据的医学检验知识图谱的领域空白,本文提出一种基于LIS构建医学检验知识图谱的方法与系统。首先通过概念层设计和实例层学习构建了包括1234个实体结点和29956条关系的医学检验知识图谱,然后将医学检验知识图谱应用于检验查询和合理检验两个方面。所构建的LIS医学检验知识图谱具有良好的知识表示结构、高速的信息查询和深层次的关系推理等优点,能提高医生的工作效率,同时也能开拓LIS系统的应用场景。To address the absence of medical examination knowledge graph based on clinical laboratory dataset of Laboratory Information System(LIS),this paper presents a method for constructing the medical examination knowledge graph.First,through the design of concept level and the learning of instance level,a medical examination knowledge graph with 1234 entity nodes and 29956 relationship edges is constructed.Then the medical examination knowledge graph is applied to examination query and reasonable examination.The constructed medical examination knowledge graph has the advantages of better knowledge representation,high-speed information query and deep relationship inference,which improves the work efficiency of doctors and enriches the application scenarios of the LIS.

关 键 词：LIS 医学检验 知识图谱 知识查询 关系推理 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]