Table 1.
The information listed in Table 1 is extracted from the Reference Manual 1200 Subject Release from the Consortium Human Connectome Project Page 189. Gender and five year-age range are available as Open Access data. Additional information about e.g., demographic status is restricted. Corresponding recordings are available in https://www.humanconnectome.org/study/hcp-young-adult.
Fig 1.
Illustration of the different reduction schemes applied to MEG signals to simulate the international 10–20 system of EEG electrodes.
Green circles indicate the positions of EEG-electrodes, and red circles closest MEG-detectors of EEG channels Fz and O2. Panel a, b, and c refer to connections scheme C1, C2, and C3 respectively. A detailed explanation is provided in the method section of the main text.
Table 2.
MEG scheme reduction to the 10–20 EEG system.
MEG detectors selected for the simulation of the international 10/20 EEG system. The first column corresponds to EEG-electrodes, the second column corresponds to C1 configuration of MEG sensors. The third column denotes the MEG sensors used in schemes C2 and C3.
Fig 2.
Illustration of the methodology applied in this study, namely the experimental data and the data processing pipelines.
A detailed description can be found in the main text.
Fig 3.
Stationary correlation pattern (SCP) in MEG recordings.
Panels a to f show average cross-correlation matrices derived from the 126 MEG detectors of a subject with ID number 105923. Averages are taken separately for different task conditions: a and b correspond to the motor task M1 and M2, c and d to the working memory task W1 and W2, e for the resting state R and f displays the cross-correlation matrix averaged over all trials of the 5 conditions. Panel g shows for each subject the topological similarity between correlation matrices averaged separately for each condition. Panel h displays the empirical cumulative probability distribution of the topological similarity between subjects separately for each condition. Finally, panel i shows the same for the correlation matrices averaged over all 5 conditions (SCP) separately for each subject.
Fig 4.
Qualitative comparison of MEG and EEG correlation matrices.
Average correlation matrices derived from MEG and EEG recordings. The upper row shows the reduced correlation matrices of MEG-recordings averaged over all 5 physiological conditions of the subject with ID number 105923. Reduction scheme C1 to C3 is used for panel a to c respectively. In the second row, cross-correlation matrices derived from EEG recordings using different reference schemes, averaged over sleep stage 2 of a whole night sleep of a clinical healthy subject (subject 10 of [10]) is shown. For panel d the global average has been chosen, for matrix e the earlobe electrode A1 has been taken and, finally, for matrix f the active electrode Cz is used as the EEG-reference. The last row displays cross-correlation matrices derived from the EEG recording of the same subject during night sleep, using the median instead of the global average as a reference scheme. From panel g to i correlation pattern is averaged over sleep stage 2, slow wave sleep, and REM sleep respectively. The order of the electrodes (from left to the right, from below to above) in each of the matrices is: Fp1, F3, C3, P3, O1, F7, T3, T5, Fz, Cz, Pz, Fp2, F4, C4, P4, O2, F8, T4, T6.
Fig 5.
Similarity of the stationary correlation pattern (SCP) between MEG and EEG recordings.
Pairwise comparison of the stationary correlation patterns obtained for the sleep EEG’s of the 10 clinically healthy subjects using the median reference and the reduced average correlation matrices obtained from the MEG recordings of the 48 subjects. The cumulative probability distributions of the Pearson coefficients are shown for the separate comparisons with the three different reduction schemes C1 to C3.
Fig 6.
Similarity of the stationary correlation pattern (SCP) and correlation matrices estimated for data windows of varying sizes.
For each length we chose randomly 15 windows of a given length. The samples of the resulting Pearson coefficients are drawn as box plots as a function of the window sizes.