Mapeamento sistemático da literatura em aprendizado de máquina para identificação de fibrilação atrial

Abstract

Atrial Fibrillation is one of the arrhythmias that most affect the elderly population, with a predisposition of several risk factors. Electrocardiograms are used to check the frequency of heart pulses and whether the internal waves are in a steady, rhythmic rhythm. Through the use of machine learning using ECGs as data, it is possible to predict and detect its occurrence. This work aims to synthesize some of the works present in the literature that used machine learning to detect heart disease. Through the studies found in the literature, it was possible to analyze that the classification methods that believed better performances were those that used deep learning.