基于大数据技术的糖尿病临床决策支持系统分析与展望  

Analysis and forecast of clinical decision support system for diabetes meHitus based on big data technique

作　　者：司家瑞[1,2] 穆得虎 孙俐[4] 乔桢[1] 杨昆[5] 

机构地区：[1]天津医科大学基础医学院,300070 [2]天津医科大学公共卫生学院,300070 [3]天津医科大学生物医学工程与技术学院,300070 [4]天津医科大学护理学院,300070 [5]天津医科大学信息中心,300070

出　　处：《国际生物医学工程杂志》2017年第3期216-220,I0004,共6页International Journal of Biomedical Engineering

基　　金：国家自然科学基金（71673199,81671054）;天津市科技计划项目（15ZCZDSY00500）

摘　　要：糖尿病是一种慢性非传染性疾病，目前只能通过长期用药和自我管理来缓解病情，无法根治。临床决策支持系统能够模拟糖尿病医疗专家诊断疾病的思维过程，向医生提供常规诊疗方案，推荐最优方案。现有的临床决策支持系统大多基于临床指南、规则、案例推理以及本体。大数据技术可获取和处理多元异构的各类数据，提供更科学的个性化诊疗方案。近年来已有基于决策树、神经网络、模糊逻辑、支持向量机、APRIORI关联规则与多维分析和时序挖掘等多种大数据处理方法应用于糖尿病的临床诊断，但其尚处于起步阶段。对基于大数据技术的糖尿病临床决策支持系统的框架进行了分析，并展望了未来的诊疗方式。Diabetes is a chronic noncommunicable disease, which is can＇t be cured, and only can be suppressed by long-term treatment and self-management. The clinical decision support system can simulate the thinking process of diabetes specialists in disease diagnosis, and can provide the regular medical treatment plans and recommend the optimal plans to doctors. Most of the existing clinical decision support systems are based on clinical guidelines, rule-based and case-based reasoning as well as ontology-based systems. The big data technology can acquire and process multiple heterogeneous data, and provide a more scientific personalized treatment plan. In recent years, a variety of big date processing methods have been applied to the clinical diagnosis of diabetes based on decision tree, neural network, fuzzy logic, support vector machine, APRIORI association rules and multidimensional analysis, and timing mining. However, these methods are still in preliminary stage. The framework of diabetes clinical decision support system based on big data technology was analyzed, and the future diagnostic and treatment methods were forecast.

关 键 词：糖尿病 大数据技术 临床决策支持系统 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论] R587.1[自动化与计算机技术—计算机科学与技术]