Fig 1.
Example of EEG pre-processing by ICA.
Time course (5 seconds, filtered between 1 and 70 Hz) and corresponding Power Spectral Density (PSD) of: First row: original EEG signals (Fp2), where the ocular artifacts are evident (vertical arrows), corresponding to the low-frequency activity below the 5 Hz in the PSD. The value of Higuchi’s Fractal Dimension (FD = 1.525) is also displayed. Second row: the estimated IC corresponding to ocular artifacts. We also show the topography of IC weights over the scalp. Third row: EEG signal after removal of the ocular artifactual IC and the FD (1.607). Noteworthy, the ocular artifacts affects FD, being the FD value of the cleaned signal higher than the FD value of the noisy signal.
Fig 2.
A. Topographies of mean values of fractal dimension in the 3 groups: young adults (<25 years of age), adults (25–66 years of age), old adults (> 66 years of age) B. EEG electrodes are indicated by a full circle. Scatter-plot of fractal dimension values over age (years) and fitting line.
Table 1.
Correlation between interhemispheric symmetry indexes of FD, Theta, Beta and Gamma band Power values between homologous electrodes, and the age.
Fig 3.
A. Scatter plot of HArS between F3/C3 and F4/C4 over age and fitting line (r = 0.47, F(1,38) = 11.13, p = 0.002) B. Scatter-plot of mean fractal dimension values of F3 and C3 (white circle) and F4 and C4 (black circle) hemisphere over age. The gray lines connect the left and right values of each subject. Fitting line for both values are displayed (dashed line for left hemisphere: r = 0.62, F(2,37) = 11.57, p<0.001; full line for right hemisphere: r = 0.63, F(2,37) = 12.39, p<0.001). C. Topographies of the Spearman’s rho coefficients of the correlation between age and relative power in beta band. Only values for which p <0.05 (FDR corrected) are displayed.