When calculating percent signal change, I calculate a baseline using 3dTcalc
When running 3dDeconvolve, I usually use a censor file, to remove bad, outlier-ish time points (e.g. due to motion, scanner glitch, etc).
What is a conceivable way to use a similar concept (e.g. 1D censor file), to specify the points i want to use for my baseline calculation? (ie i don't want to include outlier timepoints in my average. or for block designs, maybe i'd only want to use average base condition time points for my baseline)
Is there a (easy way) to do this in AFNI? (a -censor option in 3dTcalc would be sweet...)
Is there any mathematical/statistical merit in calculating the baseline differently?

Thanks