Design And Implementation Of Hospital Data Center System Based On Hadoop

With the development of the national economy, and continuously push forward the process of hospital information, clinical information systems like electronic medical records and PACS systems have been widely used, the data storage problems increasingly evident. Data security, reliability and data backup cannot be guaranteed by the traditional data storage method, and you cannot tap the potential value of the data, so data center construction put on Hospital information technology agenda. Given the traditional data center uses UNIX server with high costs, lack of computing power, data read speed is relatively slow and other defect, this paper with the current trend of computer technology, offered to build Hadoop distributed framework based on hospital data center systems.In this paper, consider the actual needs of the hospital for data storage and the characteristics of the Hadoop framework, build the hospital data center using Hadoop, HBase, Hive, Mahout, etc. which are widely used in the era of big data technology. The data center overcomes the shortcomings of traditional UNIX servers, meets the needs of the hospital, and also provides private cloud and machine learning platform dealing with big data for the hospital. This thesis completes the development of the hospital data center systems and makes structured data which is produced by HER stored in the data center with high-speed. This thesis makes the unstructured data which is produced by PACS system stored in the data center and realizes the recommendation engine algorithms on the machine learning platform.The hospital’s data center system has low implementation cost based on Hadoop, It’s a secure and reliable data storage system with strong computing ability and easy to expand. Data center systems integrates the existing hospital clinical information systems and can store structured data and unstructured data with high-speed. On the basis of the data stored centrally on the use of Hadoop cluster computing power, machine learning platform to build a large data mining information behind the data, providing doctors with clinical decision support. The data center system solves hospital data storage problems, helps the hospital making progress in information, integration and intelligence, and promotes the development of hospital information.