Dear Varina,
There is much discussion of these topics in the literature, but no single answer.
For ICA-based analysis it is not so necessary (and this is our recommended strategy).
Otherwise, you can use fslmeants in conjunction with csf and wm masks (possibly obtained using FAST on a structural image and then transformed into the fMRI space) to extract the time courses that you want.
All the best,
Mark
On 19 Dec 2013, at 16:07, Varina Wolf <[log in to unmask]> wrote:
> Hello Experts,
>
> Do the csf and wm time series need to be regressed out on an individual level in order to correct rs-fMRI filtered_func_data.nii.gz prior to group level roi-seed correlation analysis? Or can this somehow be done on a group level? If so, then what is the best way to obtain the group level csf & wm time series for regression?
>
> Thank you much!
> Varina
|