Dear SPMers,
Here is a question about modelling event duration in a an event-related
fMRI design. Pardon me for this being a naive question, but I'm a beginner
in fMRI. Each trial of my task comprises 2 different phases I wish to look
at in terms of BOLD signal: the first phase is an anticipatory period
which is signaled by a cue predicting a reward and presented for 2,5 sec,
and the second phase is the receipt of that reward (the 2 phases are
separated apart by a variable delay -i.e. a visual target- ranging from
2,5 sec to 5,5 sec). The task consists of 228 of these trials, all
randomly intermixed.
My concern is about how to model the anticipatory phase. I don't know
whether I "have the right" to model it as an event of duration > 0 sec, as
the design is event-related. From a cognitive point of view it makes sense
to view the anticipation phase as a process lasting about 2,5 sec
(duration of the cue), but from a "hemodynamic point of view" I don't know
if it
makes sense to model that event with a duration of 2,5 sec. I actually set
up the 2 models, and though I can see that the activated regions are
identical, the model with a 2,5 sec duration gives significantly higher
T-stats... But here again I'm not sure how to interpret it, especially
whether I can draw the conclusion that it actually is a better model...
Thanks a lot in advance for your help/advice (just in case the piece of
information is relevant to answer my question, I used a TR of 2,5 sec)...
Regards,
Guillaume
--
Guillaume Sescousse
Institut des Sciences Cognitives, CNRS
67 Bd Pinel, 69675 Bron
France
Tel: 00 33 (0)4 37 91 12 44
Fax : (334) 37 91 12 10
|