Detecting atrial fibrillation using statistical features of electrocardiographic interbeat interval

Abstract

This article presents a technique for detecting atrial fibrillation (AF). The technique employs electrocardiographic interbeat interval (IBI) alteration. The IBI is the time interval between individual beats of electrocardiogram. Two statistical, mean and standard deviation of IBI are examined. For the classification a support vector machine (SVM) with RBF function is applied. We have examined this technique using clinical electrocardiogram of patients with AF. The examination shows that the detection which utilizes mean and standard deviation of two consecutive segments provide the performance of 95.17%.