Dear Katharina,
the models for statistical analysis of longitudinal data use a repeated
Anova design. When you define the contrast for the comparison between
the time points, a contrast image is obtained that represents the
average differences between the time points, taking into account that
the data is intra-individual. This essentially estimates the differences
from the smoothed images (after correcting for subject effects and other
nuisance effects).
Best,
Christian
On 25 Mar 2020, at 9:02, Katharina W. wrote:
> Hi Christian,
>
> just to make sure: Is the difference of the segmented images
> calculated from the already smoothed volumes?
>
> Thank you and best wishes,
> Katharina
|