Why would you want to compare without convolving with the HRF?
If you have an a priori region you're interested in, you can just use a mask to restrict analysis to just that ROI.
________________________________________
From: SPM (Statistical Parametric Mapping) [[log in to unmask]] on behalf of An Re [[log in to unmask]]
Sent: Thursday, October 25, 2012 6:37 AM
To: [log in to unmask]
Subject: [SPM] (too?) simplistic try to determine areas of the BOLD-effect?
Hello,
I wonder if it is possible to get a rough estimation of the BOLD-effect by only calculating the correlation coefficient between a matrix which contains the time evolution of the single voxels and a design vector (as I first try I used the block design vector without convolution with the HRF)?
I use the following code fragment
for k1 = 1:size(M1,1)
for k2 = 1:size(M1,2)
for k3 = 1:size(M1,3)
C(k1,k2,k3) = sum(squeeze(X(k1,k2,k3,:)).*vec')./sum( squeeze(X(k1,k2,k3,:)).^2 );
end
end
end
and I did not get any clusters in the back part of the brain? Is there an error in reasoning?
(The normal SPM analysis of my SE-EPI images results in really low t-values...)
Feedback is appreciated :)
Andi
|