Dear SPM experts,
We have conducted an event related fMRI study (one event type, 35 events
per run, 3 functional runs per subjects). We are especially interested
in temporal differences in activation between different brain regions
(i.e. onset of activation). (During preprocessing, we performed
slice-timing.) We have built a design matrix entering the 3 runs/subject
and using 'only' the hrf function as basic function.
When plotting 'event related' responses (fitted response and PSTH), SPM
is not able to model different event onsets.
Is it better to use 'hrf + time deriv. + dispersion' to model different
event onsets and duration? What is the right plotting-option
(event/epoch-related response-fitted response and PSTH for the contrast
'effects of interest'?)
Has slice timing to be performed in this case too?
Is there any possibility to plot the average event-related time-course
for a whole cluster of activation (that approximates the raw time
course). We have thought of using the VOI function in SPM99, 'cutting'
the variable xY.u in sections around each SOA and averaging those
sections. Is that an appropriate way to do this?
Has any adjustment (i.e. for effects of interest) to be applied to the
data when performing the VOI extraction?
Thank you very much in advance
Bernhard
|