Dear colleagues,
we calculated a PPI (using the gPPI tool). We were interested in
differential connectivity between two conditions for two seed
regions
(one seed region was located in the left hemisphere, one region in
the right hemisphere (the homologous region)).
On the second level, we built a flexible factorial design with 4
within-subject conditions (seed1, condition1; seed1, condition2;
seed2,
condition1; seed2, condition2). We used this approach because it
enabled us not only to look at the results for both seeds
seperately, but also to also investigate whether both seeds had a
common connectivity pattern. Contrasts of interest were e.g. the
differences in connectivity for seed 1, for seed 2, or the
conjunction of both.
We, however, also looked at the main effect of seed. Here we found
increased connectivity to the ventricles for one seed region
compared to the other region (see attachment). I am not really
interested in this contrast but I am interested in the
explanation. When you look at the "effect of interest contrast"(see
attachment), you can see that the regions are inversely correlated
to the ventricles. Our results
show, in my opinion, that the two regions have different noise
components. Why is that? Or do
you have another explanation?
Best wishes, Andreas
--
Prof. Dr. Andreas Jansen
Laboratory for Multimodal Neuroimaging (LMN)
Department of Psychiatry and Psychotherapy
University of Marburg, Germany
Phone: +49 - (0)6421 - 5865273