Design And Implementation Of Multi-source Heterogeneous Healthcare Big Data Governance Platform

The medical and health industry is an important support for the implementation of a healthy China strategy and one of the "top ten industries" for the conversion of old and new kinetic energy in Shandong Province.The "Life Cycle Exposure and Health Status of the Whole Population" project is a project study conducted by the Shandong Provincial Health Commission on the health status and health trends of Shandong residents.The project covers a wide range of regions and populations,including 8.5 billion pieces of health and medical data in 17 prefectures and cities in Shandong Province.These health and medical data come from a wide variety of sources and cover many types of medical data,such as residents' electronic health records and basic public health,health checkups,clinical diagnosis and treatment,disease detection,and health and medical insurance.These multi-source data present the characteristics of huge data volume,wide data sources,diverse data structures,scattered data storage modes,and uneven data quality.These data greatly reduce the availability of health and medical data,it is difficult to directly analyze and mine data,and be effectively used.Therefore,it is necessary to manage health and medical data in order to develop its potential value and provide a good big data foundation for the analysis and research of medical data.Designed and implemented a multi-source heterogeneous health medical big data governance platform with the support of the Shandong Provincial Health and Health Committee's "Research on the entire life cycle health risk exposure and health status of the entire population" project,combined with the characteristics of its health medical data and existing governance issues,To implement data governance on the multi-source heterogeneous health data in Shandong Province integrated by the project.The system uses webstorm development platform,and uses Nodejs back-end framework combined with AngularJS front-end framework for system design and development.In terms of functions,the system fully takes into account the requirements of collecting,connecting and fusing various forms of data,verifying the quality of medical data,terminology specifications of medical data,and display of national health atlas,and designing multiple functional modules to meet user health and medical data governance Demand.The specific functional modules of the system mainly include 8 functional modules:medical data integration,unified medical data model,medical data quality management,medical data fusion,medical record text structure,medical health map visual display,disease population data extraction and association,and system management.In order to achieve health and medical data governance,the system establishes a unified data model to integrate medical data.At the same time,key technologies such as medical data missing completion methods and medical term normalization methods are proposed and used to check the quality of medical data and perform medical terminology review.Standardize to improve the quality of medical data.In order to make the medical data application richer and more diverse after data management,the system designed the medical record text structured and medical data visualization,and proposed and used the disease cohort generation method to extract data and correlate related information from the disease population.In the data storage and storage process,the Mysql database is used as a data storage tool to store various medical data,and data management is applied to more than 10 million pieces of health and medical data for more than 1 million patients.Health medical big data governance platform standardizes medical terminology for more than 1.89 million disease diagnosis data,more than 2.6 million medication data,more than 700,000 surgical data and more than 1.2 million inspection item data,making the health medical data completely or approximately match medical standard term In large categories,the accuracy of standardized matching is 93.4%,which realizes the standardization and standardization of medical terms and greatly improves the quality of medical data.For multi-source heterogeneous health medical big data governance platform for more than 3,000 users,it is targeted at user groups such as staff of the Shandong Health Commission,health data analysis researchers and engineers,clinical doctors and public health professionals,and graduate students.Health and medical data governance services provide convenience for users' medical data research and analysis and mining,and urge them to realize the potential value of health and medical data,and make health and medical big data more vigorous.