Biomedical Signals Analysis Using the Empirical Mode Decomposition and Parametric and non Parametric Time-Frequency Techniques

In this paper, the Empirical mode decomposition (EMD) technique is used, in a first time, to separate the artifact from the biomedical signals. The biomedical signals treated are electroencephalogram (EEG) and Electrocardiogram (ECG) for normal and abnormal patients. Analysis of EEG and ECG signals is a challenging problem due to the fact that the signals are multi-component and very non-stationary. Due to the nature of these signals, the time-frequency analysis is an important tool for representing the evolution of the frequential components of these biomedical signals over time. The time-frequency techniques used are the parametric, Periodogram and Capon methods, and non-parametric Smoothed Pseudo Wigner-Ville method. As a first step of analysis, the EMD technique is applied to eliminate the artifact from these signals after that the parametric and non-parametric time-frequency techniques are processed to the results signals to present the non-stationary multicomponent of these signals in function of time. The results show that the EMD technique associated to the Periodogram method give a good localization of the transient abnormal ECG and the energies of the EEG signals as compared to others time-frequency methods associated to the EMD technique