Hi - there are two points here:
1. I presume you have organised your contrasts so that (eg) contrast 1 in
all runs is testing for the same condition (eg "A vs E").
2. For this single-subject multi-run analysis you probably want
fixed-effects not the mixed-effects model that FEAT gives you - that's
probably why you are not seeing any activation. Have a look on the FSL
email list archives for the command-line programs needed to run
fixed-effects averaging.
Cheers.
On Mon, 28 Jun 2004, T Meindl wrote:
> Hi,
>
> we examined a patient in a single session with 5 runs and 5 different
> cognitive tasks (A,B,C,D,E)each run using a block design. Tasks were
> repeated twice a run, task sequence was altered each run.
>
> Run01: E, A, B, C, D, E, D, C, B, A
> Run02: B, E, A, D, C, C, D, E, B, A
> Run03: ...
> Run04: ...
> Run05: ...
>
> Now I'd like to look at the mean effect of each task, for example A.
> I started a First-Level analysis for each run and got good activations (for
> example Run01A.feat, Run02A.feat, Run03A.feat ...)
> Doing a Higher-level analysis with this lower-level feat directories to see
> the effect of a task over the session (in the example A) I got no
> activation at all. The matrix I used consisted of 1 group with 5 inputs,
> EV1=1 and Contrast EV1 = 1 (as described in the FSL handbook for single
> group average).
> The question is: is there a better or any other way to design the higher
> level analysis?
> Hope you can help me
> and many thanks
>
> Thomas Meindl
>
Stephen M. Smith DPhil
Associate Director, FMRIB and Analysis Research Coordinator
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|