Research On Data Mining Based On Electronic Medical Records Of Coronary Heart Disease

Objective：To explore the indicators related to patients and diseases incoronary heart disease (CHD) electronic medical records, the CHDclassification was made. Aiming to provide a reference for CHD clinicaltreatment programs, we made a further check on the correlation and theirvarious categories of association rules between the test results andmedication.Methods: Preprocess the data by using SQL database technology;cluster the clustering data by using the K-means of fast clustering; analysisthe correlation relationship between the test results and medication by theusage of association rules analysis.Results： CHD was subdivided into8new categories;59strongassociation rules between the test results and medication were mined.Conclusion：8new categories and57Strong association rules betweenthe test results and medication of CHD have an instructive significance ontargeting to develop CHD clinical treatment programs.