afni.nimh.nih.gov

Oh, and maybe add in one more censoring criterion:

-regress_censor_outliers 0.05 \

... which would censor EPI volumes that have more than 5% of the brainmask filled with outlier values at a given time point?

--pt