Advancements and Challenges in Heart disease Prediction Model: A Survey

Cardiovascular diseases continue to be a major concern in the global health forum which is estimated that they contribute to more than 60% of all yearly adult deaths in the world according to the World Health Organization. A heart’s inability to pump blood effectively results in high blood pressure, which exacerbates heart disease and raises the risk of sudden death. To meet this dire need, there is a need to come up with diagnostic systems that can be relied on to detect these diseases in the shortest time possible and with maximum accuracy. Such systems would therefore facilitate the treatment and check on persons, to ensure that they do not develop complications such as heart attacks. The use of Deep Learning (DL) and machine learning (ML) has demonstrated promise recently in this area and provides a range of advanced automated techniques for the diagnosis, treatment, and prognosis of cardiac disorders. Specifically, the developed review aims to present an overview of the related detection techniques, features extraction process, commonly used datasets, and performance indices. Particularly, it analyzes cases like Cleveland Heart Disease data, MIT-BIH Arrhythmia data, Framingham Heart Study data, and so on. The study compares approximately 25 research papers with a focus on the methodologies, difficulties, and findings in terms of computational complexity, data demands, feature constraints, and model portability. As this research aims to fill the literature gaps, it also aims to provide a generalized solution for various issues in heart disease detection and healthcare results improvement.