Preliminary Study Of Cardiac Functions Data Model By Ultrasound In Diabetic Mellitus

Objective:To study cardiac function (the left ventricular diastolic Functions) anddetect the key influencing factors and characteristic index of diabetic mellitus patientsby using data mining technologies be based on the theory of data mining getting patientsbasic information and clinical information from data database. Digging out thepotential、useful informations between the impact of factors and cardiac function. Toanalyze the trend of the various indicators and get the dynamic changes of the heartfunction. To built the data model forecasting the change of the diabetes，heartfunction、interfere the cardiac complications as soon as possible.Offering the probabledecision support for dynamically monitor、clinically treat and observe the treatment.Methods:360diabetic mellitus patients were detected by conventionalechocardiography from2012.7to2014.2. Firstly, we measured the height、weight、abdomen circumfenceand BSA of subjects;Secondly, analyzed the cardiac configurationand function parameters and to check the basic information and clinical one of thepatients (patients with or without diabetic mellitus genetic history, the level of bloodglucose、blood lipids、blood pressure, whether high-sugar、high-caloric diet, whetherhigh-protein diet, whether smoking, whether drinking and ECG changes, etc.) with theconnect of hospital information system. This as a data source to create the database，the database management tool is SQL Server2008， the data mining tools isSQLServer2008data mining module, To study the basic clinical information and leftventricular function of diabetic mellitus patients,Dig out the potential and value informations between the impact of factors and cardiac function. To analyze the trend ofthe various indicators and get the dynamic changes of the heart function. To built thedata model forecasting the change of the diabetes，heart function、interfere the cardiaccomplications as soon as possible.Offering the probable decision support fordynamically monitor、clinically treat and observe the treatment. To preliminary establishthe easy mode reflecting the change of the left ventricular diastolic function of thediabetic mellitus.Result: To establish the relational database of diabetic mellitus patients about thebasic information、clinical one and the data about cardiac configuration and function,And to convert the relational database to the transaction databases,Provided a way toanalyze the left ventricular diastolic function factors and some associated measurementindicators about the diabetic mellitus patients,We can know the cardiac function closelyrelated to genetic history、high blood pressure levels、blood glucose、blood lipids、leftatrial volume index by the association rule mining, The data mode as aided diagnosissystem was design preliminary by data mining tools.Conclusion: Data mining technology can make full use of the large amount ofmedical information, which can timely and accurately predict the cardiac function toextract the valuable diagnostic rules, and has a certain practical significance of clinicalaided diagnosis.