Have you tried to do a 2 sample t-test at the 2nd level in SPM with your groups?
You should probably use the same ROI for both groups; if you're doing it based on group activation, then the intersection of the 2. It would be even better if you had an a priori ROI (based on anatomy, or function, or whatever makes the most sense); then you could use that same region for both groups.
________________________________________
From: SPM (Statistical Parametric Mapping) [[log in to unmask]] on behalf of Mürner-Lavanchy, Ines [[log in to unmask]]
Sent: Wednesday, February 27, 2013 3:05 AM
To: [log in to unmask]
Subject: [SPM] Creating ROIs for extraction of Percent Signal Change
Dear Experts,
I would like to examine the Percent Signal Change (PSC) difference in certain areas of two groups during an fMRI task.
Normally I would create ROIs (with MarsBar) based on the whole group brain activity resulting from the SPM fMRI analysis and extract PSC.
In this case, however, I have two groups which show activations in almost the same area but not the same cluster size.
Can I create different ROIs (e.g. in the same area but with different size) for each group (based on their group activation), extract PSC and then compare PSCs to each other?
Or is it nessesary to build the same ROI for both groups (say a ROI which includes the activated areas of both groups) in order to compare PSCs?
Can you recommend papers which discuss this question?
Thank you very much for your help!
Ines
|