Dear FSLers, I have a question regarding how to removing the CSF signal from my cleaned up resting-state fMRI data (prior to running Melodic). I have a mask for CSF and a mask. After digging around JiscMail I have done the following: fslmeants -i filtered_func_data.nii.gz -m CSFmask -o CSFmean.txt fsl_glm -i datan.nii.gz -d CSF_mean.txt --out_res=filtered_func_data_res From what I understand, filtered_func_data_res.nii.gz is the data with CSF regressed (is this data demeaned, it runs from around -800 to 800). Do I then stick this data into Melodic with the pre-stats set to off? Thanks a bunch, Bryson