ECG Analysis via Machine Learning Techniques: News and Perspectives

Diseases affecting the heart have also become common day by day as a result of a bad lifestyle involving medical conditions such as diabetes and hypertension. ECG is the gold standard that is widely used for the diagnosis and prognosis of various heart conditions. However, manual inspection of ECG signals is a difficult task and can be s ubjective, time-consuming and susceptible to inter-observer variability. The extraction and segmentation of features in the ECG plays a significant r ole in the diagnosis of most heart disease. Classification s ystems are becoming increasingly popular; their main task is automatically analyze different heart diseases using machine learning and deep learning methods to improve diagnostic accuracy. The main objective of this work is to give an overview of the various machine learning approaches used to analyze the ECG useful for the implementation of Computer-Aided Diagnosis (CAD) allowing to support the diagnosis of the main heart diseases such as myocardial infarction (heart attack), differentiate arrhythmias (heart rate changes), hypertrophy (increased thickness of the heart muscle) and enlargement of the heart. The different methods are presented with reference to the application context, to qualitative and qualitative parameters, to the adopted algorithms and to the obtained results.