Dear SPMers,
I have searched for a similar problem in the mailing list but couldn't find clear solution.
I have two active conditions A and B with implicit baseline. For DCM-GLM
i've specified three regressors:-
1. A+B
2. A only
3. B only
This has been done so as to be able to define (A+B) as main driving input in DCM and A & B to be modulatory inputs.
I've then defined three contrasts:- 1 0 0, 0 1 0 and 0 0 1 to define VOIs specific to (A+B), A and B respectively. strangely, it shows deactivations for first contrast(1 0 0) in the same regions where A(0 1 0) and B (0 0 1) show activations. Also, activations for (0 1 0) and (0 0 1) are exactly same for 25 subjects.
The design matrix shows all parameters to be estimable (see attached image design matrix.jpg).
When i plotted the fitted responses (predicted), SPM shows inverted form (-ve) of regressor1 to be fitted to data. (see attached image fitted response.jpg)
Can you please help me to find where i'm going wrong.
Regards
Monika Grewal
|