Automatic Detection and Prediction of Paroxysmal Atrial Fibrillation based on Analyzing ECG Signal Feature Classification Methods

Paroxysmal atrial fibrillation, a really life threatening disease, is the result of irregular and repeated depolarization of the atria. In this paper, an automatic detection and prediction of this critical disease is performed by the use of three groups of features extracted from different parts of ECG signals and classified by KNN, MLP and Bayes optimal classifiers. Finally, the health status of more than 90% of cases are diagnosed correctly and also it is possible to detect an ECG record far from the PAF onset from the one which is immediately before PAF onset in more than 70% cases.