Hi SPMers,
We have collected some fMRI data that we believe includes a SENSE
artifact in the form of "streaks" of activation that are perpendicular
to the plan of acquisition. We have taken some steps to reduce SENSE
(e.g., reduction of coverage area), but have not been able to
eliminate the artifacts completely.
Our question is, what are the likely consequences of these artifacts
in the data already collected? While we recognize that their presence
is clearly sub-optimal, we are unsure how to gauge their severity. The
streaks do not appear in our primary ROIs, nor do they appear in the
same place for each subject. Is it reasonable to assume that these
artifacts will generally average out in an RFX model, or will they
substantially increase our risk for Type I or Type II error, or both?
Any guidance would be very helpful.
Many Thanks,
Michael
Michael Treadway, MA
Department of Psychology
Vanderbilt University
301 Wilson Hall
111 South 21st Avenue
Nashville, TN 37203
610.909.2565
[log in to unmask]
|