There are strange patterns of negative correlations in the correlation maps generated by the connectome pipeline.
(1) #1 likely cause is scrubbing not taken into consideration.
(2) assuming the problem is indeed scrubbing, then confound regression strategy for non-scrubbed volumes should be changed so that no artefacts are induced even if non-scrubbed volumes are used.
There are strange patterns of negative correlations in the correlation maps generated by the connectome pipeline. (1) #1 likely cause is scrubbing not taken into consideration. (2) assuming the problem is indeed scrubbing, then confound regression strategy for non-scrubbed volumes should be changed so that no artefacts are induced even if non-scrubbed volumes are used.