For the pre-processing steps, MATLAB (Mathworks, v2016a) and EEGLAB toolbox (Delorme and Makeig, 2004) were used. The closed eyes segments were selected and the electrodes were computationally located on the scalp using the dipfit toolbox (Oostenveld et al., 2011) using the standard 10–20 electrode model. Afterward, the data were bandpass-filtered from 1 to 70 Hz and bandstop-filtered from 45 to 55 Hz using the pop_firws function in MATLAB, with a filter order of 2, and the Kaiser window parameter beta was estimated using a maximum passband ripple of 0.001; the data were downsampled to 200 Hz. Afterward, the data were divided into 1-s epochs, and the epochs with excessive noise or artifacts as judged by visual inspection were removed. Spherical interpolation was applied for channels with excessive noise, drift, or bad connection. The EEG had to have less than or equal to three electrodes with excessive artifact; otherwise, the EEG was excluded from the analysis. Afterward, the EEGs were re-referenced to average reference, and independent component analysis was performed using the extended infomax algorithm (Lee et al., 1999) for each
