Predicting the Myocardial Infarction from Predictive Analytics Through Supervised Machine Learning

The pattern variations in the ECG signal image, i.e., the ECG report image, help to identify the myocardial infarction. This paper enlightens us regarding very effective preprocessing techniques to get structured and very clean data for the feature selection and feature extraction process. This study extracts unique features from ECG graph images. The different machine-learning methods make the diagnosis very simple and easy within less time. Particularly, variation in the output of electrodes 2nd and 3rd will intimate heart attack, and the remaining electrodes will also show variations, respectively. The authors obtained 14 features and applied them to the present classifiers giving more efficient results, among which GBC showed 98.79% of test accuracy. Finally, the gradient boosting classifier was able to identify different heart attacks effectively. © 2023, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.