Autoregressive models of electrocardiographic signal contaminated with motion artifacts: Benchmark for biomedical signal processing studies

Abstract

Motion artifacts are one of the biggest challenges in analysis of electrocardiographic signals in outpatients, different signals processing techniques have been developed to significantly reduce these artifacts, however, there is not a standard signal that consider artifacts and allows to compare the proposed algorithms to reduce these, as is done in other areas such as acoustic or seismic. In this paper, a mathematical model for the representation of the electrocardiographic signal with motion artifacts is proposed; Fourier analysis and adjustment parameters are shown to mathematically represent an electrocardiography lead. Identification of motion artifacts in ambulatory electrocardiography was performed by autoregressive models. Additionally, the obtained signal from the mathematical model and its performance are shown.