A Review on Heart Diseases Using Machine Learning and Deep Learning Techniques

Heart disease is one of today’s major issues, as well as one of the main causes of mortality globally. Electrocardiogram (ECG) and patient data may be used to diagnose cardiac illness in its early stages, as shown by recent advancements in machine learning (ML), deep learning (DL) application. Many scholars and practitioners have uncovered numerous data level and algorithm level solutions throughout the years. An exhaustive literature review is presented here to reveal the difficulties posed by unbalanced data in heart disease forecasts and to offer a wider perspective on the available knowledge. Using 93 articles of reference that we had acquired between 2018 and 2023 from respectable journals, we conducted a meta-analysis. An extensive analysis of 30 literature references has been done, taking into account the kind of cardiac ailment, methods, programmes, and results. Our research showed that existing approaches have a few problems that have yet to be resolved when working with datasets, which ultimately reduces their real usefulness and efficacy. ML and DL methods are utilised to enhance data-driven decision-making for cardiac disease detection. For content analysis of 93 articles’ metadata, 30 articles about the diagnosis of heart disease were chosen. The research was primarily concerned with the models’ performance, as well as other issues including the machine learning and deep learning’s interpretability and explicability.