Functional Connectivity Analysis with AAL Atlas

The SPM Marsbar toolbox (version 0.44, http://marsbar.sourceforge.net/) was used to obtain the parcellation of the subjects’ fMRI volumes in $N$ = 90 regions of interest (ROIs) of the Automated Anatomical Labeling (AAL) atlas^{33 (link)}, representing the nodes for the functional connectivity (FC) analysis. We selected all AAL ROIs except from the ones located in the cerebellum and vermis, which were not covered by the fMRI volumes in some or all subjects.
For each subject, we extracted the mean BOLD time series of the voxels within each ROI (node). Using Matlab in-house scripts, instantaneous statistical dependencies among ROIs were assessed by computing the Pearson correlation coefficients between the BOLD time series of each pair of ROIs, resulting in a $N\times N$ FC adjacency matrix for each participant, whose elements represent the pairwise cross-correlation between the BOLD time series of the corresponding ROIs. Only the functional connections corresponding to significant Pearson correlation values (p < 0.05) were considered, by setting to zero the non-significant ones. The resulting subject-level FC matrices (either weighted or binarized using an arbitrary positive 0.5 threshold) were further analysed to extract FC features of interest.