EEG recording and data analysis
EEG data were recorded in an eyes-open resting condition with the
BioSemi system (Biosemi Inc., Amsterdam, the Netherlands), using 128
Ag/AgCl electrodes set and analyzed with MATLAB software (version R2019)
running EEGLAB10 functions and proprietary scripts.
Raw data were bandpass filtred (0.5-50 Hz) and manually examined for
artifacts. Contaminated segments and bad channels were removed from the
analysis and then interpolated. In the next step, EEG data underwent an
Independent Component Analysis (ICA) to exclude remaining artifacts. The
PSD-EEG was calculated using the MATLAB pwelch function, with a Hamming
window of 512 samples and 50% overlap in six frequency bands: delta
(1-4 Hz), theta (4-8 Hz), lower alpha (8-10 Hz), upper alpha (10-13 Hz),
beta (13-30 Hz) and gamma (30-40 Hz)3. Individual
maximal peaks in the theta-alpha band were detected using the
“findpeaks” function in the 4-13 Hz frequency range. For all channels
and frequency bands, we constructed topographic maps of significant
inter-group differences. These statistical maps revealed clusters of
significant electrodes over midline frontal and left frontal lobe. Based
on the polar coordination system, and according to literature
data,11–13 we matched the electrodes to three Regions
of Interest (ROIs): the supplementary motor area (SMA − C14, C18, C19,
C20, C27), the prefrontal motor cortex (PFC − D04, D05, D06, C31, C32),
and the primary motor cortex (M1 − D10, D11, D12, D19, D20). Finally,
from each crucial region, we selected one representative electrode, i.e.
an electrode with the most significant inter-group difference: C21
(SMA), D05 (PFC) and D10 (M1) (Figure 1A) for further analysis.