医疗大数据特征挖掘及重大突发疾病早期预警  

Characteristics Mining of Large-Scale Physiological Signal and Early Warning of Major Diseases

作　　者：滕琪[1,2] 樊小毛[1] 何晨光[1] 李烨[1] 卢东昕[2] 

机构地区：[1]中国科学院深圳先进技术研究院,深圳518055 [2]江西省南昌大学软件学院,南昌330047

出　　处：《网络新媒体技术》2014年第1期50-54,共5页Network New Media Technology

基　　金：中国科学院战略性先导科技专项"面向感知中国的新一代信息技术研究"(XDA06000000)课题"鲲鹏大数据系统及产业化示范"(XDA06010500)资助

摘　　要：通过运用新兴的尖端云计算技术,设计研发了健康云平台。采用分布式云存储技术存储大规模的异构多模态生理信号数据,将数据挖掘模型(L1-逻辑斯蒂)整合至MapReduce框架中,快速挖掘用户的健康信息以及重大疾病高风险因子,让用户能实时地了解自己的身体状况,同时对用户的异常状况给出预警信息,并通知其前往医院就医,实现了对重大突发疾病的早期预警。In this paper,we design and develop a health cloud platform using state-of-the-art cloud computing technology. We use distributed cloud storage technology to store the large-scale heterogeneous multimode physiological signals and the date mining models（ L1-logistic） integrated into the MapReduce to analyze user health information and high-risk factors of major diseases quickly, which make users to be aware of their physical condition immediately. Users are given the early warning according to their abnormal condition and informed to hospital for medical treatment at the same time. It realized the early warning of major diseases.

关 键 词：云计算 生理信号 数据挖掘 

分 类 号：TP311.52[自动化与计算机技术—计算机软件与理论]