hi Irina
> 1) Am I right that all the onsets indexes have to be reduced by one, so if the event comes
> together with the first pulse, it has the onset=0, with the 5th pulse - onset=4, and so on?
Right - time 0 (either scan = 0, or 0 seconds) is the onset of the first scan.
> 2)I removed the first 5 volumes of my session, since I didn't want them to be processed
> with others. So I only want to use the volumes from XXXX_006.nii to XXXX_400.nii for
> my model. What if I still have the events corresponding to the pulses 1:5 in my onsets
> vector? Do I need to remove these events? Do I need to change the numbering of all
> events so that onset "0" corresponds to the 6th pulse?
You're exactly right again: The first scan you keep (in your case
XXX_006.nii) is time 0; so everything else is relative to that. If
you have events that occur before that, there shouldn't be a problem
assigning them a negative onset time.
Hope this helps!
Jonathan
|