Thank you very much for the responses.
So just to confirm that I'm doing the right thing:
fsl_glm -i (nifti file) -d (.txt file with the confounds) -o nifti_betas.nii --out_res=nifit_confoundCorrected.nii
I use the nifti_confoundCorrected as input for my ICA analysis in MELODIC. The results are not completely concordant with my expectations. When I look at the nifti_confoundCorrected with FSL view, a lot of volumes look completely white or completely black. Should I edit my confounds before performing confound correction (whitening and demeaning) before applying the confound correction? I cannot find a clear answer in literature about that.
Thank you,
Debby
|