Hi all,
I processed my task-related fMRI data (block-design) using SPM8 and used the ART toolbox to determine potential outliers. I then ran GLM and gPPI analyses.
A reviewer is arguing that using this approach for artefact denoising, the global signal is included as nuisance regressor, which has shown to introduce negative correlations. Is this correct?
He suggested me to use the aCompCor approach instead. As far as I understand it, these 2 approaches are very different and do not address the same issues. Could anyone clarify?
Thanks for your time.
Best regards,
Yann
|