Dear all,
As I didn’t receive any response to my previous question, I post it here again. Could someone give me any advice on the issues below?
In a simple block design experiment I have two conditions let’s say Condition 1 (C1) and Condition 2 (C2) and a Baseline (B)
I have created contrasts of the t-stat like
C1-B
C2-B
I have, therefore, C1- and C2-specific activations for each participant. What I want to do is to find regions activated by both C1 and C2. At the first level I just chose the two contrasts and I ran a conjunction analysis. However, this analysis does not produce any contrast images to be used at the second level analysis. When I do the random effect analysis I can create SPMs for the individual contrasts but not for conjunction of contrasts. Does anyone know how to do this 2d level conjunction analysis? An alternative would be to use some weighting like [1 1 -2], with the first two numbers corresponding to C1 and C2, and the third number to Baseline. Would this approach be more liberal than a conjunction analysis?
I was also thinking of using inclusive masking and as I am interested in areas specific to C1 and C2 I want to also implement an exclusive masking of the two contrasts. Again this procedure does not produce any contrast images to be used at the second level. I was thinking of using the imcalc function to calculate a new image of let’s say contrast C1-B with inclusive masking of contrast C2-B e.g. i1.*(i2>0.05) as well as exclusive masking e.g. i1.*(abs(i2)<0.05), for each participant. Then I am planning on using the new images for the second level analysis. Is this the correct approach? Do the expressions for exclusive and inclusive masking make any sense to you?
Thanks in advance
Alex
|