图书情报知识 ›› 2019, Vol. 0 ›› Issue (1): 109-118.doi: 10.13366/j.dik.2019.01.109

• 情报、信息与共享 • 上一篇    下一篇

基于扩展疾病本体的电子病历大数据组织研究

陆泉,江超,陈静   

  • 出版日期:2019-01-10 发布日期:2019-01-10

Research on Big Electromic Medical Record Data Organization Based on Extended Disease Ontology

  • Online:2019-01-10 Published:2019-01-10

摘要:

[目的/意义] 电子病历大数据作为主要的医疗健康大数据,具有较高的研究与应用价值,但其多源异构特性阻碍了其有效利用,需要针对性研究其组织问题。[研究设计/方法] 本研究从电子病历大数据与疾病知识体系的联系出发,提出了一个基于扩展疾病本体的电子病历大数据组织模型框架,通过多源知识聚合进行疾病本体知识体系扩展,进而基于扩展后的疾病本体知识体系进行电子病历大数据映射,实现了电子病历大数据的知识描述与组织。[结论/发现] 利用维基百科与MIMIC数据集进行了实验验证,实验表明模型框架与方法行之有效。[创新/价值] 本研究可以有效支持多源异构电子病历大数据的知识组织与多样化的精准数据访问,拓展了医疗健康大数据的组织与利用途径,有助于提升医疗健康大数据利用率。

关键词: 电子病历, 知识组织, 疾病本体, 本体扩展, 多源知识聚合, 数据映射

Abstract:

[Purpose/Significance] As the important medical big data, big electronic medical record(EMR) data has high value for research and application. However, it's difficult to use medical big EMR data effectively due to its multisource and heterogeneous characteristics. Therefore, it is necessary to study the organization of big EMR data for its characteristics. [Design/Methodology] This paper focuses on the natural relations between big EMR data and the disease knowledge system, and then proposes a framework of big EMR data organization model based on extended disease ontology(DO), multisource knowledge aggregation and medical big data mapping, which realizes integrated organization of disease knowledge system and big EMR data. [Findings/Conclusion] This paper conducts an empirical study by using Wikipedia and MIMIC dataset in order to validate the framework and method of the mode, which is found to be feasible and effective. [Originality/Value] The research can provide effective support for knowledge organization of big EMR data and multiknowledgedimensional precise accessibility, which expands the organization and application of medical big data, and contributes to improve the utilization of medical big data in hospital.

Key words: Electronic Medical Record(EMR), Knowledge organization, Disease ontology(DO), Ontology expansion, Multi-source knowledge aggregation, Data mapping