Hi Andrew,

I would expect the effect of eye movement on motion parameters
to be negligable. Even supposing the eyes did not have the
lower intensities that they usually do have, volume registration
should be dominated by the relatively huge mass that is the brain.

Under what circumstances have you seen eye movement affect volume
registration?

---

Note the way afni_proc.py creates an EPI mask. It uses 3dAutomask
per run, and then takes the union of the results. Note also that
by default it does not even apply this mask, as that often leads to
trouble in group analysis, since each subject will have different
mask coverage. And most people seem to then forget to intersect
these subject masks. And those that remember are often depressed
by the volume of brain that had been removed from the analysis.

So now we do not mask in single-subject space. That step is saved
for a group mask (from the anat or EPI group template).

- rick