Atrial Fibrillation and Premature Contraction Classification Using Convolutional Neural Network

Irregular rhythm is the common classification of arrhythmia. The abnormality of heart is described as arrhythmia. This abnormality or functioning of heart is recorded and interpret as Electrocardiogram (ECG) signals. These signals have appropriate information regarding type of arrhythmia. It is tedious to analyze manually ECG signals because of their non-linearity and intricacy. In this write up, a distinct calculative AI based methodology used for ECG signal which is Deep learning. The proposed methodology achieves with ECG data inputs which is transfigured into ECG images by mapping single ECG beat as individual 96x96 image and these images passed to the CNN. It learn the key features of used Atrial Fibrillation (AFB) arrhythmia database which is used for classification. These autonomous features directly learn by the CNN model with having different layers from ECG signal. This method will remove the human error to specify the diagnostic feature while working with big data. The case study and results show evident and approve the capability of the learned features for atrial fibrillation and flutter. The CNN classifier is used for classification which gives good results.