Dealing with noise and physiological artifacts in human EEG recordings: empirical mode methods
Anastasiya E. Runnova, Vadim Grubov, Marina V. Khramova, Alexander E. Hramov · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2017
In the paper we propose the new method for removing noise and physiological artifacts in human EEG recordings based on empirical mode decomposition (Hilbert-Huang transform). As physiological artifacts we consider specific oscillatory patterns that cause problems during EEG analysis and can be detected with additional signals recorded simultaneously with EEG (ECG, EMG, EOG, etc.) We introduce the algorithm of the proposed method with steps including empirical mode decomposition of EEG signal, choosing of empirical modes with artifacts, removing these empirical modes and reconstructing of initial EEG signal. We show the efficiency of the method on the example of filtration of human EEG signal from eye-moving artifacts.