were also included as covariates in the design matrix. Finally, in order to assess positive relations of the dual-regressed time series we implemented contrast masking, i.e. group compar- ison results were masked by group average maps (voxels that fell within either DLB or control group average maps, thresholded at p - value < 0.05, familywise error (FWE) correction for multiple compar- ison using threshold free cluster enhancement (TFCE)). No statistical correction for multiple RSNs was implemented.