The Research And Implementation Of Electronic Medical Record Data Quality Analysis System Based On Data Mining Techniques

The medical data is informationization essential foundation for each big Medical establishment, However along with the various hospitals application system data quantity's sudden expansion, the new application has as well as between the application mutual conformity unceasingly, each kind of application is increasing the information unceasingly every day toward the database, the data also increases by the exponential order speed, data quantity and the complex degree are unprecedented . The data quality question becomes day by day prominent, these questions mainly display in the data not correctly, the data incomplete, the data inconsistent and so on aspects. The quality inferior data already became affects the hospital superintendent to carry on correct policy-making the important attribute, therefore the data quality control will certainly to become in the medical profession informationization advancement an essential link.The electronic medical record is the digitized hospital core and the foundation, is the digitized hospital diagnosing and treating information resource and the service platform. Electronic medical record sources of information in hospital information system's each subsystem, like doctor workstation, nurse workstation, Laboratory Information Management System (LIS), Picture Archiving and Communication Systems (PACS), as well as other Clinical Information System (CIS) and Hospital Information System (HIS). These subsystems its concerned patient's basic information and diagnosing and treating information submission for electronic medical record, thus forms patient's electronic medical record. The electronic medical record's content originates in provides the patient for it to diagnose and treat the information the subsystem, its data comes from the different time one by one, different hardware platform, different operating system's data integration. Because in inside many kinds of data pool's data will have"the dirty data"massively , therefore the data must undergo certain processing, will use the data extract tool, through will obtain the data the characteristic and the data integrity analysis, will realize the data quality control.In view of the above question, this article has done certain research, the work and the achievement is as follows:1. Studied and designs one to have the flexible electronic medical record data quality analysis ETL construction. This construction divides the rank based on the rule the data quality analysis method, through joins"the electronic medical record data qualitative index collection"and"the quality analysis rule storehouse"carries on the data quality problem analysis to the source data, through buffer database, data extract filtration and data quality classification control function, but enhances the data quality analysis effectively the efficiency and the rate of accuracy.2. Developed and has realized a new ETL construction electronic medical record analysis system. This system may use the new construction fully the characteristic, has to the ETL flow's performance enhances greatly. At the same time, through to the data quality regular storehouse's refinement, raised system's analysis efficiency effectively.3. Realized the data quality subject multi-dimensional OLAP analysis demonstration. Through quality analysis's subject establishment multi-dimensional anatomic model, and may carry on each kind of OLAP operation, many kinds of way demonstration data quality analysis result.