Dear SPMers,
I have some questions regarding seed-based functional connectivity analyses on task-related fMRI. My task is based on a blocked design and includes two conditions (A and B) and rest. I am interested in how the time series of my region of maximum activation in condition A>B (i.e. a T contrast of 1 -1) is functionally coupled to the rest of the brain.
I have performed my 1st level analysis and included my two conditions (A and B) as well as my 6 motion regressors into the GLM. I have then looked at my contrast of interest (A>B) and extracted the time series of my activation maximum (VOI).
My questions are the following:
1) I understand that one possibility to perform a seed-based functional connectivity analysis is to use the extracted time series (VOI) as a single regressor for another GLM for the same subject (by using the realigned, normalised and smoothed images), however, without specifying any conditions, but using one contrast only (+1 in the contrast manager, so it would include conditions A, B, and rest) and without adding motion regressors again. Is this correct?
2) Since I am definitely interested in task-related functional connectivity changes for the contrast A>B, I was wondering if I could I perform this kind of analysis including specification of my conditions and contrast A>B? How could I add my time series as a regressor of interest? Or could this question only be addressed by a PPI analysis?
3) For a PPI: I would create the ppi interaction term by using the previously extracted (condition-specific) VOI. My contrast of interest is 1 -1 (i.e. condition A > condition B), so my contrast weights for the PPI would be:
1 1 1
2 1 -1
Is this correct?
Any comments would be much appreciated
Many thanks
|