Analysis and systematization of noise arising by long-term recording of ECG signal

In this paper presents analysis and revealing features electrocardiographic noise arising long-term ECG signal recording in the condition of free human activity. Analysis and systematization electrocardiographic noise conducted with consider statistical and spectral characteristics of the signal and noise. Based on the analysis revealing problems overlapping frequency noise spectra with ECG signal and non-Gaussian noise. To solve the problems revealed is proposed to use nonlinear threshold filters together with the method of empirical mode decomposition of the ECG signal. It is assumed that this approach will solve the revealed problems and improve the noise resistance processing of ECG signals for further derivation of diagnostic findings in the frequency-time domain.