Mark Gamalo wrote:
> SPM Masters,
> We are doing an study on what areas of the brain are correlated with the
> mean arterial pressure (MAP) while working on a Stroop task. We have two
> conditions Incongruent and Congruent and we have 4 blocks of 240 scans.
> The first design matrix includes the effects of the two conditions while
> being modulated by MAP. We want to ask three main questions: What areas
> are activated due to condition? What areas are correlated with MAP? What
> areas are correlated with MAP taking out the effect of task?
>
> To answer these questions we set up the following contrasts:
> A: 1 0 -1 0 0 0 0 0 0 0 repeated 4 times for the 4 blocks; this would give
> the areas activated due to task (Incongruent - Congruent)
> B: 0 1 0 -1 0 0 0 0 0 0 repeated 4 times for the 4 blocks; this would
> crudely be the effect of task weighted by the MAP if I understand it
> correctly. The problem with this contrast is that it is not estimable at
> all. Any remarks on this?
>
> The third question would have been easy if B is estimable because it can
> just be masked by the result of A.
>
> Note: the six zeros following the first 4 digits are covariates of no-
> interest.
>
> So we tried another design this time ignoring the task. So it is a simple
> regression of the signal onto MAP. In this way, we can ask the question
> which areas are correlated with MAP. But we could not determine which
> areas are correlated with task. So another design needs to be done by
> taking only the condition without modulating it with MAP. This then would
> give areas correlated to task. We were thinking that a conjunction
> analysis can be done to mask out effects of task but we are not sure how
> to do this. Can anybody give suggestions?
>
I'm not sure I understand your design but if you
set up regressors 1 and 2 to demarcate congruent (C)
and incongruent (I) blocks and
regressors (3) and (4) to be (C) and (I) parametrically
modulated with MAP then (i) the main effect of the stroop task can be assessed with
the F-contrast
[1 -1 0 0]
(ii) the main effect of MAP during Stroop tasks can assessed with the
F contrast
[0 0 1 0; 0 0 0 1]
and (iii) the interaction between task and MAP can be assessed with the
F-contrast
[0 0 1 -1]
Hope this helps :-)
Best, Will.
> Thanks a lot!
>
> Mark Gamalo
>
>
>
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7478
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|