Hi Alfonso, 
I have 6 patients with deep brain stimulators in the nucleus accumbens. In each, I performed a block-design fMRI task, where they viewed neutral and 'rewarding' pictures. I performed the fMRI run with the DBS activated, and also with the DBS OFF (randomized order). So, 2 fMRI runs for each subject. In addition to performing a task-based fMRI activation study (which I did using FSL-FEAT), I'd like to do a connectivity analysis using CONN, but I'm struggling with the contrasts/conditions. The contrast of interest for each subject is 'rewarding' pictures vs 'neutral' pictures (1 -1), but I'd really like to know how this is affected by turning the DBS ON.
For each subject I've created conditions with their custom timings (Neutral-OFF, Neutral-ON, Rewarding-OFF, Rewarding-ON). Is it possible to use these 4 conditions, in a seed-to-voxel analysis, to model the change in connectivity to an ROI, for the effect of looking at a 'rewarding' image vs a neutral image? Since I'm using a block design, I've been working in the weighted-GLM mode, because I read somewhere on this forum that this may be appropriate for block-design tasks. I think this would be like a 3rd or 4th level analysis in FSL...
Thanks in advance for any help!
Ben