Automated Indicator of Atrial Fibrillations Risk Using Machine Learning

Abstract

Atrial fibrillation (AF) is a multifactorial arrhythmia linked to common cardiovascular diseases associated with classical cardiovascular risk factors. Although awareness and enhanced detection of AF have improved over the past decade as the incidence and prevalence of AF increases, the trends of applying machine learning techniques in the diagnosis still lack precision. This paper includes a review of the literature on the most common ML algorithms implemented to detect the risk of atrial fibrillation (AF). In this review, studies of AF published in academic journals between 2016 and 2021 will be evaluated to analyze the application strategies, the adopted techniques in the field, and the investigated research issues related to applying ML algorithms. The results of the review indicated that great progress has been made in terms of applying ML algorithms have been procured; however large variation can be seen between studies and countries. It was also found that most studies focused on further developing the identification, prevention, and risk separation of AF using machine learning techniques. Additionally, studies have indicated that technological and methodological advances in AF diagnosis made great progress throughout the years helping to prevent future prevention. The CNN algorithm has the main impact for the automated atrial fibrillation indicator using machine learning, as it has proven to be the most effective in terms of detection. Consequently, this research paper will review the most common ML algorithms in the research field, their characteristics, datasets, tools, and techniques used to distinguish atrial fibrillation risks using ML algorithms.

Keywords: Machine learning algorithms, Scientific computing, Atrial fibrillation, Machine learning, Market research, Cardiovascular diseases, Computational intelligence