Detection of Arrhythmia from ECG Signal Using Bat Algorithm-Based Deep Neural Network

Heart disease is one of the deathliest diseases in the world. As the ECG data is widely available, many machine learning and deep learning models have been developed for the automatic Arrhythmia diagnosis system. In this study, BiLSTM with the Bat algorithm has been proposed for the classification of the Normal ECG signal and the Arrhythmia ECG signal. It can detect the abnormality of the heart rhythm. The novelty of this study is the feature selection using the Bat algorithm that enhanced the performance of the classification model. The performance of this method is compared with the existing methods and it gives better performance than the other existing methods. With the proposed method, an accuracy of 99.75% has been achieved using tenfold cross-validation. It can help the risk patient in the early detection of the abnormality of the heart rhythm.