Detection of Attention-to-Rest Transition from EEG Signals with the Help of Empirical Mode Decomposition

Abstract

In this paper, an empirical mode decomposition (EMD) scheme is applied to analyze the steady-state visually evoked potentials (SSVEP) in electroencephalogram (EEG). Based on EMD method, the oscillatory activities of the decomposed SSVEP signal are analyzed. It is observed that the 6^th IMF showed the features of the attention-to-rest transition response. In other words, high powers are observed instantly after the volunteer turns from an attentively focusing stage into an unfocused attention stage. Having made the point that the 6^th IMF of the SSVEP signals corresponds to very low frequency (0.5 – 2 Hz), this drives us to look into that frequency range of the SSVEP signal. All of this reflects that a very low frequency seems to occur during the attention-to-rest transitions. Experiments are performed with different people. The result shows that the attention-to-rest transition can be detected with an accuracy of 82.6%.