Data Driven Regional Health Decision Support System Design Research

[Objectives] Regional population health information platform is the center of regional health management and decision.The data of the platform comes from the three basic databases of regional population health information.To build a regional health management decision support system that conforms to the “big data” is necessary for exploring the hidden patterns and features of the complex data and can provide scientific and reasonable basis for decision makers.[Methods] This paper analyzed the tendency of the development in health big-data by literature research,expert consulting,comparative analysis and other research methods.At the same time,this paper defined some related theories and concepts of the health big-data,and built the framework of the Decision Supported System.The paper summarized the bigdata processing method and procedure of these heterogeneous,structured,and non-structured big data which comes from different health systems.Furtherly,this paper put forwarded a parallel computing solutions based on the Spark,and use UML to design the data processing,data preprocessing and data type management modules.Lastly,this paper analyzed the application scene by SparkR.[Results] This paper uses technology of big data platform construction and residents' health data,and does the research below:1.The service requirements analysis of Regional Health Management Decision Support System.According to the requirements of Regional Health Management Decision Support System,this paper analyzed the function module and business module.The function module contains public health,medical service,basic health,Medical care management and medicine management.The business module contains query and regulation,performance evaluation and forecast planning.2.The study of key technologies for big data of regional decision support system.This paper analyzed the key technologies needed for functional modules,and chose the combination of the key technologies.These technologies consist of HDFS,NoSQL,data mining technology and Visualization technology,etc.3.According to the specific business needs of the Medical decision support system for big data,this paper realized the overall design,logic design and model base design.At the same time,this paper designed the list of unstructured data storage based on HDFS,data preprocessing module based on ETL and data storage of model Spark Sub-modules based on internal storage.Above all,this paper designed the decision model by Spring and Java DAO interface.4.After analyzing the requirement of Decision Support System,this paper designed the application scenarios by SparkR data analyzing platform.[Conclusions] This study is based on the regional population health information platform for health care data,and completed the framework design of regional health management decision support system.Regional health decision support system based on healthcare data used big data technology and platform to provide real-time,visual basis for decision-making for decision makers.Compared to the past,the system is more suited to the needs of big data era decision.