Aprimoramento da detecção do complexo QRS em exames de eletrocardiograma na presença de anomalias

Abstract

Cardiovascular diseases are a set of disorders of the heart and blood vessels, including coronary artery disease, cardiovascular disease, cardiac arrests, and other conditions. According to the World Health Organization, four out of five deaths from cardiovascular diseases occur prematurely in people under 70 years of age. The diagnosis of these diseases is usually made through an electrocardiogram (ECG), which records the heart’s electrical activity. The standard ECGs has 12 leads, each representing graphically the differences in potential in different parts of the body, allowing for the analysis of the cardiac cycle and the identification of anomalies. The most important waveforms in the ECG are, P, Q, R, S and T, with the set of them being called the QRS complex, representing ventricular depolarization. Five characteristic anomalies of the QRS complex denote cardiac pathologies. This research was based on the detection of the QRS complex in 12-lead ECGs using convolutional neural networks, improving on previous work by considering the presence of these anomalies.