Frequency, time-frequency and wavelet analysis of ECG signal

Abstract

The analysis and segmentation of an electrocardiogram (ECG) signal is a hard and difficult task due to its artifacts, noise and form. In this paper; we analyze the ECG signal in Frequency, applying Fourier transform, autoregressive moving average (ARMA), multiple signal classifications (MUSIC), as well as the short-term Fourier transform STFT, Choi-Williams and Wigner-Ville for time frequency analysis and wavelet analysis. The analysis has been done in modified lead II (MLII) of ECGs data files of the MIT-BIH database, obtaining better results of segmentation of QRS complex by wavelet analysis.