Dirk,
This paper provides a general method for determining contrast
weights:
Using larger dimensional signal subspaces to increase sensitivity in
fMRI time series analyses Volume 17, Issue 1 , Pages 13 - 16
Hope this helps,
Eric
Quoting Dirk den Ouden <[log in to unmask]>:
> Dear all,
>
> I'm playing around with modeling the 'true' hemodynamic response
> for
> individual subjects (after stroke), using FIR. This works fine
> for main
> effects, when I give the correct weights to my time bins.
> However, I'm not
> sure how to model contrasts subtracting one condition from
> another. Should I
> simply give the negative weights to the condition I am
> subtracting?
>
> Say, I have 10 time bins, which I can give a certain weight to
> resemble the
> hrf that I have measured with a different method:
> 0.25 0.4 0.6 0.8 0.9 0.85 0.7 0.5 0.2 -0.1 (I'm just making up
> the numbers)
>
> And my conditions are A and B, in an FIR model with 10 time bins.
> Would my
> t-contrast subtracting B from A be:
>
> 0.25 0.4 0.6 0.8 0.9 0.85 0.7 0.5 0.2 -0.1 -0.25 -0.4 -0.6 -0.8
> -0.9 -0.85
> -0.7 -0.5 -0.2 0.1
>
> ?
>
> This way, the contrast adds up to zero, as it should, but it
> still seems odd
> to me - it seems to entail a very specific hypothesis about how
> voxels
> 'respond' to condition B ...
>
> Any comments, or references to helpful FIR work, are welcome!
> thanks,
>
> Dirk
> ************************************
>
> Dr. Dirk-Bart den Ouden, Northwestern University
>
>
>
> Aphasia & Neurolinguistics Research Laboratory
>
> Dept. of Communication Sciences & Disorders
>
> 2240 Campus Drive, Evanston IL 60208-3066
>
> Phone (lab): 1-847-467-7591
>
> ************************************
>
|