Hi, we are trying to use Melodic to remove motion components from our data
and we have run into a question. After we run melodic we correlate the
output components with the motion correction parameters output from the
motion correction algorithm in Feat. This would presumably identify those
components that are due to motion. We then remove these components and
re-run Feat. The issue we run into is that the output from this second Feat
has much lower snr, and the cope maps seem to be much noisier as well. Is
this what we should expect from our results, or is our procedure flawed? We
expected removing the motion components to increase our snr. Thanks for
your help!
Jeff
|