Dear Gonzalo,
I'm not sure why you need to skip the first two volumes of each period.
This is certainly not a normal thing to need to do.
If you do have a very good reason for needing to do this, then the
best approach is to include separate EVs (regressors) of no interest
for each time point that you want to ignore (or "skip"). For example,
if
you wanted to ignore timepoints 0, 1, 10, 11, 20, 21, 30 and 31 then you
would include 8 separate EVs. The first EV would have a one at
timepoint
0 and zeros for all other timepoints. The second would have a one at
timepoint 1 and zeros for all others. The third would have a one at
timepoint 10 and zeros for all others. And so on. These are easy to
make with custom EVs or by using the square wave input with appropriate
"skip" and "stop after" settings. It is also important that you turn
off both convolution and temporal derivative (but leave temporal
filtering on).
If you do this then you will effectively remove all of the signal
contributed from these timepoints, but without introducing the nasty
effects which would occur if you just removed these timepoints from
the data.
Before you do go ahead and do this, however, please think about why
you would do this in case it is not necessary.
All the best,
Mark
On 29 Jul 2006, at 18:48, Gonzalo Rojas Costa wrote:
> Hi:
>
> I have an fMRI study with TR = 4 seconds, 10 temporal dynamics per
> off-
> on period (40 seconds duration each off-on period)... I must skip the
> first two dynamics of each off-on period to process the study... How
> can I
> do it with FSL ?...
>
> Sincerely,
>
>
> Gonzalo Rojas Costa
> Service of Neuroradiology
> Institute of Neurosurgery Dr. Asenjo
|