2.6. Brain connectivity analysis

The functional connectivity analysis were carried out in the CONN toolbox. The preprocessed BOLD signals were submitted to evaluate global brain connectivity (GBC, GBCi=1n∑jrij2, where r_ij is the correlation coefficient between voxels i and j, n = #voxels inside the brain), which is a graph theory‐based connectivity metric assessing network centrality at each voxel (Martuzzi et al., 2011). The GBC is a whole‐brain voxel‐based measure of connectivity, addresses qualitatively different question about brain connectivity than seed‐based analysis. Aberrant GBC might suggest disturbed information processing from a brain region to other brain areas.

The seed‐based FC maps were obtained by computing the Fisher‐transformed correlation coefficients between the average BOLD signals in rACC and the signals in all other brain voxels. The seed rACC was defined as a 6‐mm‐diameter sphere centered on a point with MNI coordinates (x = 0, y = 38, z = 4), based on the coordinate from a previous placebo study of open‐label antidepressant medication treatment (Sikora et al., 2016). The group rACC seed‐based FC map was created by performing a random effects one‐sample t‐test across all participants (Figure ​(Figure1),1), with age, gender and mean framewise displacement (FD) as nuisance covariates.

One‐sample t‐tests results of the baseline rACC seed‐based FC (p < .05, FWE correction at the cluster level with a voxel‐level threshold of p < .001 uncorrected). Two confound regression strategies were used: (a) aCompcor and (b) global signal regression