Dear fsl experts,
I am struggling with a feat analyis. Eugene helped me with it last week and it is still a problem. I have 2 groups of 12 subjects each, and each subject completed 3 5 mintue fmri scans. Eacn scan involved viewing different types of emotion faces, and the first-level analysis was to examine amygdala connectivity across the 5-minute scan. I want to look at group effect, scan effect, and interactions.
I have 72 inputs, each a result of a 1st-level connectivity analysis. The first 36 are group 1, first 12 of those scan 1.
For the EVs, I have set up:
- 1 column modeling group (top half 1, bottom half -1)
- 3 columns modeling scan (combination of 1,-1 and 0), same pattern for each group
- 3 columns modeling the interaction (product of group and scan columns).
- 6 sets of 11 columns modeling the difference between subject 1 and the other 11 subjects for each group and for each scan.
For the contrasts, I have 7 contrasts, 7 F tests, where C1: EV-1=, all others 0, C2:EV2=1, all others 0, etc.
This model was impossible to load into the gui, so I had to write a script to get it into a .fsf file. Now it's all in, but I have the error: Problem Processing the model: Terminate called after throwing an instance of "NEWMAT::Program Exception.'
Any suggestions? Does the model I described seem right for what I want to do?
Thanks, Katie
|