Hi Dr. Beckmann and other FSLers,
I often find that a MELODIC time course component is a mixture of motion
effects and a design component. This is made abundantly clear when I include
the motion correction parameter vectors as explanatory variables in the
MELODIC analysis! Nevertheless, the "IC time course" is often systematically
different from the "full model fit" time course. To avoid throwing the baby
out with the bath water I would prefer to remove only the full model fit. So
I have three questions. Is this advisable? If so, can MELODIC presently do
it? If not, would a file that is composed of the full model fit vectors, but
is otherwise similar to the "melodic_Tmodes" files do the trick?
Regards,
Kent Conover
|