Hi,
On 28 Jan 2007, at 19:32, Vina Goghari wrote:
> Hello,
> I am starting use FSL for my analyses for the first time. I am
> conducting an
> experiment with jittered event-related design, therefore I am using
> 3-column
> format to analyze the data. I understand that the data needs to be
> organized, by <onset> <duration> <1 or 0, depending on whether it
> is of
> interest>.
Not quite. For a given EV and associated 3-column timing text file
you only model one kind of event, and you don't need to model the
gaps in between events. Hence you would normally only have 1s in the
final column. If you have a second event type, that you are not
interested in, that should be modelled by a separate EV and
associated timing file, again with 1s in the 3rd column. Please see
the FEAT manul (the "details" section) for details on this.
> In the datafile created by my stimulus program, my first MRI scan
> and object
> onset corresponds to an onset time of 363 ms (and not 0 ms),
> because the
> script has instruction screens etc that take up the first 363 seconds.
> Since, I'm using the onsets from the script, is that a problem that
> the
> first event and MRI scan chronicle it occuring at 363 ms and not 0ms?
For all timing values, including the onset timings in the 3-column
timings file, these have to be relative to the exact start of
acquisition of the first volume. Hence if the scanning started at
363ms, you need to subtract 363 from all your timing onset values.
> Also I am using a jittered trial design with an average trial
> length of 8
> seconds. Does duration refer to duration of stimulus (which is 650
> ms) or
> duration of that specific trial? If it is the latter, I have a
> compound
> trial type with a first stimulus and jittered timing of 300-400 sec
> then a
> second stimulus and then jittered timing of 270-370 sec.
The onset and duration timings in the timing file should be correct
for each inidividual stimulus, i.e. if the stimulus durations vary
then so should the durations in the timings file. If you have to
events making up a trial you may want to model them as two different
conditions, i.e. again you may need two separate EVs and timing files.
Cheers, Steve.
------------------------------------------------------------------------
---
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
------------------------------------------------------------------------
---
|