Automated Prediction of Spontaneous Termination of Paroxysmal Atrial Fibrillation Using Support Vector Machine

摘要

Based on our previous work, we develop a classifier using support vector machine to differentiate paroxysmal atrial fibrillation (AF) from sustained AF and propose an automatic scheme to predict the spontaneous termination of paroxysmal AF. Experimental studies over the challenge database of Physionet/CinC 2004 show that our method is able to achieve high classification accuracy and is fairly reliable. This in turn justifies our previous work.