Hi Rick,
Thanks for the quick reply!
The reason I ask is:
1) I am analyzing resting-state data for the first time and I want to make sure I'm not missing anything.
2) If I understand correctly, these methods differ slightly from what Power et al (2014) implemented for their own approach?
They seem to interpolate the censored time-points to perform frequency filtering on them. I was confused by these differences, and trying to determine best practices.
3) Parts of my analysis is performed outside afni (in matlab), so I may have to implement censoring myself. It would probably be easiest for me to censor via regression, so I wanted to make sure that this was not substantially different from your approach (or if there were particular reasons not to do it that way). Alternatively, perhaps it would be most fruitful to censor in AFNI and take the resulting residuals to the next step for further analyses.
Thanks again,
Andrew
P.S. When you said that censoring is done in 3dDeconvolve the same way since 1998, you were referring to the time-point deletion method? I thought one approach might be preferred because it takes less time to compute or for other non-results related reasons.