Machine learning approach for ECG analysis and predicting different heart diseases

Abstract

In the modern world, there have been some revolutionary advancement in the field of medical science and research and this is no different for electrocardiogram. Electrocardiogram (also abbreviated as ECG) illustrates the electrical activity of one’s heart over a period of time. Over the years, number of people suffering from heart disease have increased to some extent. Therefore, in our research, we aim to design a model using supervised machine learning that can find anomalies in one’s ECG report by analyzing it. We have applied six supervised machine learning algorithms to distinguish between normal and abnormal ECG. In addition, we used them to predict the chances of a patient suffering from a certain disease. We divided our data set into two parts. 75 percent data in one group for training the model and rest 25 percent data in another group for testing. To avoid any kind of anomalies or repetitions, Cross Validation and Random Train-Test Split was used to obtain an answer as accurate as possible. We have compared the results with each other for a better understanding.