Hi all,

I'm trying to import longitudinal data that was preprocessed in fMRIPrep.  there are 2 scan session done a year apart.  At each scan there are 2 runs of a continuous performance task with 5 different stimuli types that last 2 seconds each with a 1 second fixation point between each stimulus.  These stimuli are random events that we have the onset times for.  The problem that I'm running into is that when I use the "Import fMRIPrep Dataset" tool, it imports BOTH scan sessions into 1 subject number with 4 task sessions.  

In the past, we've always split the first and second scan into 2 different subject numbers, then each subject would have 2 sessions (one for each run of the task).  In each session, there would be 5 different conditions: square, circle, positive pics, neutral pics, and negative pics.  The first and second run are different from each other with different onset times for the stimuli.

However, now I have 1 subject with the first 2 sessions being the first 2 runs in the baseline scan and sessions 3 and 4 being the two runs in the 1-year scan.  Plus I have the 5 conditions to input with 2 different onset times to input for each.  Most of the comparisons will use the square as the "baseline" (so square-neutral pic v. square-negative pic).  Also, I need BOTH runs of each condition to be included.  So you would end up with (session 3 + session 4) - (session 1 + session 2) for the conditions to be compared. Then, when you add 1-year v baseline to that, it gets really complicated to set up in the results - to the point where I have no idea how to even start.

How would I do these comparisons?  Or would it be easier to set up the project in a different way?  I tried "merging" run 1 squares with run 2 squares, but I don't know how the merging works in CONN, so I'm not sure if it's doing it correctly or if I need to adjust it somehow for it to work correctly.

Thanks.  I know this explanation may be a bit confusing, so please feel free to ask me to explain something better.

Wade