Electrocardiogram (ECG) is used for detection of a number of heart abnormalities. In this project, we have utilized a two lead ECG in order to improve the ECG signal accuracy applying artificial neural network (ANN). In order to extract features, current methods commonly rely on advanced signal processing of the main lead II (MLII). Although MLII is obtained easily, it does not include a complete description of the electrical activity of the heart. Therefore, in order have a 2D view of the electrical activity of the heart we have used a precordial lead, V1, from the Standard 12-lead ECG system. The designed method enables the user to process the inputs from any ECG measuring device that are commercially available. In this study input from Fitbit was used to distinguish normal signal from the abnormal signal.