When running resting state analyses, when you are correlated a seed with the rest of the brain, I am guessing it is best to extract the average timecourse for the seed using the unsmoothed file, to avoid spillage from other nearby brain regions. And then, correlate that timecourse with the rest of the brain (the errts file).
However, the unsmoothed data (vol reg) won’t be censored for motion (which would be problematic)?
I know the CONN toolbox makes use of the unsmooted data to extract the average timecourse of the seed.
To do what you want, just skip the blur step in afni_proc.py,
and then apply the blur to the errts result that the seed time
series will be correlated with.
Note that the volreg data would not have gone through the
“cleaning” regression model, either. Individual time series
could be passed through it (just using the X-matrix), but it
seems like less work to save the blurring for later.
This is not uncommon, and of course depends on the blur size
and how the seeds are chosen. Blur sizes are usually not too
large, so unless the seed is at the edge of a region, it might
not make a big difference.
rick
The
National Institute of Mental Health (NIMH) is part of the National Institutes of
Health (NIH), a component of the U.S. Department of Health and Human
Services.