Hi Jessica, Yes, this should work just fine. Use fslsplit, then fslmerge after fsl_glm. All the best, Anderson On 14/01/2014 11:50, Jessica Jesser wrote: > Dear FSLers, > > we'd like to regress out the CSF signal from a rfMRI dataset using fsl_glm as described in https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1306&L=fsl&D=0&1=fsl&9=A&I=-3&J=on&d=No+Match%3BMatch%3BMatches&z=4&P=565712 > Unfortunately our dataset is quite large (more than 800 timepoints) and while processing the dataset fsl_glm crashes because of an "out of memory" error (even if reserving 10GB of RAM for that process). > Would it be possible to split the dataset spatially (not temporal), use fsl_glm on every single slice and merge all those slices back again? Or does GLM use the information in neighbour voxels (resp. neighbour slices)? > > Thank you! > Cheers, > Jessica