afni.nimh.nih.gov

I second that vote for looking at the raw EPI data to see whether it contains uniformly high early values.

It might also be worth noting that this is also a feature afni_proc.py will try to check for and warn users about; it plops any potential warning in an output text file called "out.pre_ss_warn.txt", and this datum is also displayed (and potentially flagged) and the QC HTML that afni_proc.py creates.

But again, looking at the data initially is certainly the best way to start.

--pt