An exhaustive review of machine and deep learning based diagnosis of heart diseases

In comparison to other diseases, the number of deaths on Heart Disease (HD) is the highest across the globe. The trend of death due to HD is still rising which has become a constant source of concern amongst the human beings. The researchers and doctors are putting tremendous efforts to save the life from HD. It is observed from the literature that a large number of researchers are currently carrying out their research work in various aspects of HD. Among those the early detection and diagnosis of HD are currently the focus area of research. Appropriate, reliable, accurate, robust and affordable HD detection scheme is the ultimate goal for saving the lives of the people. In this research, articles on HD detection and diagnosis published in recent past have been collected and critically analysed. The outcome of the analysis is presented in various tabular forms for easy understanding and further use. The paper would provide a thorough knowledge on standard data source on HD, the feature extraction, selection and reduction methods and Machine Learning (ML) and Deep Learning (DL) based classification schemes. The categorization of published articles and the various performance measures employed have been presented which would develop interest amongst new researchers working in the area of detection or classification of HD. The best performing technique in each category of has been listed. The research challenges and future scope of work are also provided to facilitate further research work in this promising area.