Dear SPM users,
I use a fast event-related fMRI design with two events A and B randomly distributed. The SOA varies between 1.5 – 2.5 seconds with a TR=2s. If I have well understood, I need to take account of the nonlinear (saturation effect) by using Volterra interaction during the fMRI model specification in SPM. However, I have not understood how to use them.
- If I want to compare A vs. B taking account of the nonlinear effect, should I include the interactions in my contrast or should I use a classical contrast [1 -1] on the main condition ? If I need to include them in the contrast, how should I use them?
- What is the effect on the betas estimation of my conditions (A and B) if regressors of Volterra interactions are included in the model?
- What is the meaning of activation maps for the Volterra interaction regressors?
Thanks for your help!!
Michaël
|