基于Hadoop的医疗健康数据管理系统研究与设计

发布时间：2018-05-25 06:48

本文选题：医疗健康 + Android　；参考：《广西师范大学》2017年硕士论文

【摘要】：随着我国移动互联网、物联网、云计算、可穿戴设备等新技术的发展,以及惠及全民的健康信息服务和智慧医疗服务的提出,推动了健康大数据的应用。面对海量的医疗数据,高效地存储和快速地处理数据成为当前主要需求。传统的数据库对海量级的医疗数据面临着存储应接不暇、成本居高不下、计算能力无法企及等问题。针对这些问题,本文利用Hadoop架构实现对海量数据的分布式存储和处理,结合集群特性实现低成本、高效、可扩展的数据处理。本文结合医疗健康数据的特性,提出基于Hadoop的医疗健康数据管理系统,系统分为两大模块:基于Android的医疗健康管理系统客户端和基于Hadoop的医疗健康数据管理系统服务器端。基于Android的医疗健康管理系统客户端,是以Android智能手机为依托,设计移动互联的健康管理APP。该软件包括用户注册登录、心率检测、体重测量、运动测量、数据统计,健康建议六大部分,能实现生理参数的测量,改进运动测量算法,分析测量结果和提供健康建议等功能。基于Hadoop的医疗健康数据管理系统服务器端,利用Hadoop全新技术对数据进行快速存储和管理。本文搭建4台机器的集群中心,考虑到医疗数据需要频繁实时写入读取的特性,在Hadoop集群中心搭建HBase数据库,取代HDFS,并重新对数据格式进行设计,从而实现数据库HBase满足存储医疗数据。同时,使用并行计算模型MapReduce对医疗健康数据进行分析处理。此外,改进HBase多条件查询方法,将HBase与Solr结合实现多条件查询,提高查询效率。最后,设计并模拟对医疗健康数据管理系统的客户端和服务端的系统性能和可靠性测试,验证结果表明系统能够达到需求的功能设计和可靠性能,Android各个功能模块测试通过,MapReduce性能测试和HBase数据库写入测试与预期一致,HBase和Solr相结合查询数据时间大大缩短。因此,基于Hadoop的医疗健康数据管理系统能够更好的满足医疗健康数据要求。
[Abstract]:With the development of mobile Internet, Internet of things, cloud computing, wearable devices and other new technologies in China, as well as health information services and intelligent medical services that benefit the whole people, the application of health big data has been promoted. In the face of massive medical data, efficient storage and fast processing of data has become the main demand. The traditional database is facing the problems of overflowing storage, high cost and unreachable computing power for massive medical data. To solve these problems, this paper uses Hadoop architecture to realize the distributed storage and processing of massive data, and combines the characteristics of cluster to achieve low-cost, efficient and scalable data processing. Based on the characteristics of medical and health data, a medical and health data management system based on Hadoop is proposed in this paper. The system is divided into two modules: the client of medical and health management system based on Android and the server of health data management system based on Hadoop. The client of the medical health management system based on Android is based on the Android smart phone to design the mobile connected health management app. The software includes user registration, heart rate detection, body weight measurement, exercise measurement, data statistics, health advice six parts, can achieve the measurement of physiological parameters, improve the motion measurement algorithm, Analyze the results and provide health advice. The medical and health data management system server based on Hadoop uses the new technology of Hadoop to store and manage the data quickly. In this paper, the cluster center of 4 machines is built. Considering the frequent and real-time writing and reading of medical data, the HBase database is built in the Hadoop cluster center to replace the HDFS, and the data format is redesigned. In order to achieve the database HBase to meet the storage of medical data. At the same time, the parallel computing model MapReduce is used to analyze and process the health data. In addition, we improve the method of HBase multi-condition query, combine HBase and Solr to realize multi-condition query, and improve the efficiency of query. Finally, design and simulate the system performance and reliability test of the client and server of the medical and health data management system. The verification results show that the system can meet the requirements of functional design and reliability. The time of querying data through MapReduce performance test and HBase database writing test is greatly shortened with the combination of HBase and Solr. Therefore, the medical and health data management system based on Hadoop can better meet the requirements of medical and health data.
【学位授予单位】：广西师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.13;TP311.52

【参考文献】