Dear Experts,
I have a question about resting-state data analysis.
I know that there are some debates on the usage of global signal
regression on the seed-based resting-state data analysis since it
might introduce false anti-correlated networks. I was told that global
signal regression is same as global mean (it is not suggested now by
SPM and replaced by grand mean). My understanding is the global
signal regression is to remove the global signal effects ( regress
out), but the global mean is just to scale the mean and the
difference of the mean of every scan(time point) is still preserved.
So they are kind of different. Is my understanding right? anybody
could elaborate a little more if I am wrong?
Thanks for any opinions,
Xiaoying
|