We are using VBM in a study of gray and white matter differences in persons
with Williams Syndrome. The results of the GM analysis agree in part with
our hypotheses concerning reductions in patients relative to controls.
However we also find a number of areas of unexpected gray matter increases
in patients. We have observed in our statistical analyses of morphometric
data from other image analysis software that eliminating patients with
greater motion artifact produces results more consistent with our
hypotheses, but of course reduces power.
How sensitive is VBM to motion, and do ways exist to eliminate suspected
artefactual results?
Thanks.
Ken Rando
|