Hi Zhang,
It is hard to know for sure from the activation maps alone, but the images look like motion artifacts
to me, especially the "stripes" in the sagittal view. A possible cause is the spin history effect when
there is cross-slice motion for an interleaved scan sequence (Friston, et al, 1996).
My artifact repair software ( http://cibsr.stanford.edu/tools/ArtRepair.htm ) may help clear up the
activation map, as long as the subject motion is not totallly correlated with the task. Try the Artifact
Repair button to repair the images after preprocessing, and then run the repaired images in
Estimation.
Good luck,
Paul
|