Arrhythmia Classification Using Machine Learning Techniques on ECG Data

Gauri Rajendra Shende, Dr. Ayesha Siddiqui

Abstract: Heart arrhythmia disorder is considered one of the leading causes of death globally. It is crucial to identify any irregular activity in the heart at an early stage for proper treatment of the disorder. ECG signals are used to measure the electrical activity of the heart and give significant information about heart diseases. On the other hand, interpreting the ECG signals manually requires a lot of time, particularly when there is a high amount of data to be processed. This process may result in human errors that may delay the proper identification of the disease. Machine learning algorithms used in heart arrhythmia classification using ECG signals form the basis of research work. There are certain processes involved such as processing of ECG signals, feature extraction and applying machine learning techniques that allow identification of various types of arrhythmias. There exist different machine learning methods that may be applied during the classification of heart arrhythmias such as SVM, Random Forest and ANN. The publicly accessible datasets of the ECG signal will be used to evaluate the effectiveness of the proposed technique, and the results will be analyzed using the following parameters ? accuracy, precision, recall, and F1 measure. In our studies, it was found out that machine learning methods can offer better heart arrhythmia diagnosis and also yield faster results compared to traditional techniques. The study helps to develop intelligent healthcare systems which can help medical practitioners diagnose cardiovascular diseases at an early stage.

Keywords: ECG, Heart Arrhythmia Detection, Machine Learning, Classification, Artificial Intelligence, Healthcare Analytics