Fig. 8.
Intrinsic effective connectivity of the default mode network shown for each of the four sessions for 10 participants (A) and cross-session similarities of the baseline intrinsic effective connectivity A0 among 30 participants for several temporal basis sets (B); 1) a stationary model (A0stationary, mean ± standard deviation 0.234 ± 0.110), 2) a simple average model (A0avg, 0.575 ± 0.060), 3) a DCT (up to 6th component) model (A0DCT6, 0.577 ± 0.060), 4) a PCA (up to 6th principal components) model (A0PCA6, 0.588 ± 0.059) and 5) a fPCA (up to 6th functional components) model (A0fPCA6, 0.577 ± 0.060). The upper lines between models in (B) indicate significant between-group difference (p < 0.001, Bonferroni corrected). The mean cross-session similarities (among 30 participants) are displayed for the three temporal bases sets we used (C). Here the first component corresponds to the baseline effective connectivity A0 at the second-level.