信息资源管理学报 ›› 2017, Vol. 7 ›› Issue (4): 29-37.doi: 10.13365/j.jirm.2017.04.029

• 专题-医疗健康大数据挖掘研究 • 上一篇    下一篇

MIMIC-III电子病历数据集及其挖掘研究

陈静 李保萍   

  • 收稿日期:2017-06-23 出版日期:2017-10-26 发布日期:2017-10-26
  • 作者简介:陈静,女,副教授,研究方向为信息组织与信息检索;李保萍(通讯作者),女,研究生,研究方向为信息分析,Email:1640351499@qq.com。
  • 基金资助:

    本文系教育部人文社会科学重点研究基地重大项目“大数据资源的挖掘与服务研究——面向医疗健康领域”、湖北省高校省级教学研究项目“信息管理类‘知识主题-课程’体系网络构建研究”(2016078)的成果之一。

Research on MIMIC-III Electronic Medical Record Dataset and Its Mining

Chen Jing Li Baoping   

  • Received:2017-06-23 Online:2017-10-26 Published:2017-10-26

摘要:

为了解美国典型开源医疗数据库-重症监护室电子病例数据集(MIMIC)内容及其研究利用情况,本文系统调查梳理MIMIC数据集的内容,从发文数量、发文国家机构、研究主题及研究方法四个维度,对Web of Science中的有关研究文献进行文献计量与内容剖析。结果表明,对MIMIC数据集的研究利用处于上升时期,但研究深度与广度不足;我国高校与医疗公司应加强对其研究利用;研究主题侧重ICU病人预后与死亡率预测,应扩展研究主题;挖掘方法偏好通用机器学习和统计分析方法,应加强针对性挖掘方法研究。

关键词: MIMIC, 电子病历, 数据挖掘, 主题分布, 重症加强护理病房(ICU)

Abstract:

In order to understand the content and the research and utilization of a typical and open source medical database in the United States-the Medical Information Mart for Intensive Care, also called MIMIC, this paper systematically investigates the content of the MIMIC dataset and analyzes the literatures about the research of the MIMIC dataset in the Web of Science, respectively from number of documents, nations, institutions, research topics and research methods. It is found that the research and utilization of MIMIC dataset are in the ascending period, and its research depth and breadth are insufficient. Universities and medical companies in China should strengthen their researches and utilization on MIMIC dataset. The research topics focus on the prediction of prognosis and mortality of ICU patients, and researchers may explore more research topics. Data mining methods mainly focus on the general machine learning and statistical analysis methods, while mining methods with stronger pertinence need to be proposed.

Key words: MIMIC, Electronic medical record, Data mining, Topic distribution, Intensive care unit(ICU)

中图分类号: