Hi Alberto,
I think this is wrong. Any difference you found in dual regression is a
difference of coactivation within a particular component (network).
Before performing dual regression, you have to check in your GICA which
are the components (netoworks) of potential interest for your study.
Best
Angela
> dear FSL experts.
>
> I performed a TC-GICA + dual regression analysis on one group of subjects
> looking for a correlation among their RSN activities and some subjects
> covariate values. well, I cannot interpretate one thing.
> considering one component, my results are often limited to some voxels,
> that is, in my analysis, the concept of network has disappeared, and i
> find differences in voxels. may be this correct?
> consequently, when i output my results, is there a difference between a
> significant voxel encountered in IC2 or IC30 ??? do they are equally
> important ? where has the concept of component ordering gone ?
> are my results maps composed by all the significant voxel, regardless of
> the component they come from ???
> so, can i simply create one image adding all the significant voxels from
> all the component???
> thanks in advance
> Alberto
>
>
|