A systematic review on cardiovascular disease detection and classification

• This systematic review analyzes 50 research works based on cardiovascular disease detection. • Specifically, the review analyzes the different relevant works and identifies the research gaps in the literature. • Ultimately, the findings of the review assist the researchers to develop innovative and advanced technique for CVD detection. Cardiovascular Disease (CVD) is a life-threatening syndrome and the primary cause of many fatalities worldwide. Specifically, cardiac disease impairs blood vessel function and can result in coronary artery infections, which can ultimately cause mortality. Hence, there is a demand for a reliable, accurate, and feasible system to diagnose cardiac disease promptly and assist cardiologists in treating patients appropriately. However, the most significant challenge in heart disease prediction lies in identifying the behavioral patterns of CVD in the large volume of data as well as the time complexity that affects the detection efficiency. Meanwhile, the development of Artificial Intelligence (AI) techniques including the Machine learning (ML) and Deep Learning(DL) assists in the extensive and complex medical data analysis to recognize patterns in which the disease occurs and convert them as structural data for predicting the disease. Specifically, this systematic review compares 50 research publications, with an emphasis on the approaches, challenges, and findings regarding computational complexity, data requirements, feature limitations, and model portability. This systematic review analyzes the selected publications rigorously and assists in the identification of gaps in the literature making it helpful for researchers to create and implement in clinical settings, mainly on datasets pertaining to heart disease. Finally, the future prospects for CVD risk assessment and prediction are provided in this review contributes to develop more appropriate proposals for the future development of AI-based techniques.