Diagnosis of coronary arteries stenosis using data mining.

摘要: Cardiovascular diseases are one of the most common that cause a large number deaths each year. Coronary Artery Disease (CAD) is type these worldwide and main reason heart attacks. Thus early diagnosis CAD very essential an important field medical studies. Many methods used to diagnose so far. These reduce cost deaths. But few studies examined stenosis vessel separately. Determination stenosed coronary artery when significant ECG abnormality exists not difficult task. Moreover, among patients. The aim this study find way for specifying lesioned there enough changes only based on risk factors, physical examination & paraclinic data. Therefore,  a new data set was which has no missing value includes  effective features like Function Class, Dyspnea, Q Wave, ST Elevation, Depression Tinversion. were collected from 303 random visitor Tehran’s Shaheed Rajaei Cardiovascular, Medical Research Center, in 2011 fall 2012 winter. They processed with C4.5, NaA¯ve Bayes, k-nearest neighbour algorithm (KNN) algorithms their accuracy measured by tenfold cross validation. In best method reached 74.20 5.51% Left Anterior Descending (LAD), 63.76 9.73% Circumflex (LCX) 68.33 6.90% Right (RCA). found too.