Research On Health Records Data Warehousing And Data Mining

Information systems have been widely used in various medical and health institutions, with the development of information technology in China, especially in the 21st century. However, due to the lack of a unified organization and design, lack of interconnection between systems, data sharing is difficult to achieve. Particularly electronic health records, there is a serious "information island" and"Information chimney"phenomenon. In order to change this situation, in May 2009, the Ministry of Health issued"based on health records Platform construction of the Regional Health Information Guide (trial)"to establish regional health information within a unified service platform to meet the needs of users at different levels.This paper focuses on the construction of regional health information platform for the needs of health records data warehouse ,system studied on the related technology of data warehouse, architecture and implementation method. Introduced on-line analytical processing (referred to as OLAP) multidimensional data analysis methods, described the concept of data mining technology and principles. This paper studied on the data mining process and implementation steps, analyzed the multidimensional data model design of health record data warehouse. This paper realized data mining on health records using OLAP functions and association rules within SQL Server 2008 database, obtained the expected results.Major completed works include:1. Focused on the design and construction of the health records data warehouse. Combed the business process and data characteristics of the health records, divided the 5 business topics and 31 child domains. Proposed the design of the health records data warehouse conceptual model and logical model, as well as health records data cube design.2. Using SQL Server 2008 database system, created fact table and dimension table to store health records data. Designed and implemented the health records data cube,completed the physical implementation process of a health record data warehouse.3. Using OLAP technology, conducted multi-dimensional data mining for the cause of death of a city residents, reached a conclusion that the chronic non-communicable diseases were the main causes of death in the city.4. Described and analyzed the Apriori algorithm for association rule mining and applied to analyze the causes of birth defects. Mining out of 5 rules which meet specific support and credibility, as well as consistent with clinical experience.