Preliminary Study On Ultrasonic Data Model Of Cardiac Geometry In Overweight And Obese Patients

Objective: To explore the basic clinical information of overweight and obese patients and the conventional ultrasonic measurement of cardiac geometry and to reveal the potential and valuable information about the relationship between the influence factors and the change of cardiac structure by using data mining techniques.Then analyze the change rules of each index to find the dynamic changes of the cardiac structure and to provide the basis for clinical diagnosis and treatment.Methods: 182 patients(including 74 cases with overweight,49 cases with obese patients,59 cases with normal weight)in our hospital or medical examination from October 2015 to October 2016 were selected as the research object.Patients with congenital heart disease,cardiomyopathy,valvular heart disease,severe arrhythmias,and so on were excluded,patients with hypertension,diabetes and other chronic diseases that have been shown to have a significant impact on the cardiovascular system were also excluded.Obtain basic information of patients from HIS and LIS,such as blood pressure,blood glucose,blood lipids,liver and kidney function,T3,T4,TSH,whether smoking or drinking,ECG abnormalities and other basic information.Height and weight were measured,obesity duration were inquired and recorded.Conventional cardiac index,such as systolic and diastolic atrial,ventricular diameter,atrial area,left ventricular mass,atrial,ventricular volume(Simpson),left and right ventricular outflow tract diameter,interventricular septum,left ventricular posterior wall,right ventricular anterior wall thickness,aorta and pulmonary artery diameter etc,were measured.The original data was discretized and transformed into the form of transaction database suitable for mining.Use the association rules(Apriori algorithm)in data mining to analyze the ultrasonic data and related indexes of the cardiac structure and the influencing factors.After that,extracted the effective association rules and eleced the characteristics and related factors of cardiac structure and its changes.Result: Establish an association rule model of cardiac geometry in overweight and obese patients and then analyse cardiac structural parameters and related factors.The results show that the left atrial diameter,left atrial area,left atrial volume,left ventricular mass weight tended to change in overweight and obese patients,the change of cardiac structure is closely related to age,obesity degree and obesity duration.Conclusion: Data mining technology can make full use of the large amount of medical information,which can timely and accurately predict the cardiac configuration by extracting valuable diagnostic rules.It can provide reference for further data research and provide basis for clinical diagnosis as well.