> For starting the PPI pipeline I should not use 3dDetrend initially but instead use 3dSynthesize on
> my input dataset to making a dataset of my baselines/regressors of no interested then use 3dcalc
> to make a timeseries with those removed? I would believe this covers the detrending as my polorts
> were removed as effects of no interests in 3dSynthesize.
The processing route of 3dSynthesize+3dcalc is slightly more preferable for two reasons: 1) it provides a more accurate removal for the drifting effects; 2) it also allows for removing other confounding effects such as head motion.
Gang
