Concatenating our runs together leads to some huge (sometimes over 2 Gig) datasets. We have been running into some problems with files this large (namely a 2 GB file size limit on our Linux). Is there anyway for 3dDeconvolve to accept multiple input BRIKs in a multi-run experiment rather than physically concatenating the runs together into one larger dataset? If not, does anyone have any suggestions on getting around a file size limit?