Hi, I am attempting to use spm_hrf parameters estimated from a PSTH taken
from an ROI. I would like to use those parameters to run another analysis,
to see which areas correlates with the ROI.
The problem I have is that I have 2 conditions in the data, call them A and
B. The A and B PSTHs have different shapes, so it would be nice to be able
to specify these two different regressors using different spm_hrf
parameters. However, I don't think there is a way to do this.
The archives contains this message
http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind03&L=spm&D=0&I=-1&P=24378
which works around that for the temporal derivative. But in this case it's
a bit more complicated since B is not just a time shift of A.
One possible workaround is to create two design matrices with 2 different
hrf parameters, [A1 B1] and [A2 B2], and then splice [A1 B2], into SPM.xX.X.
Are there any problems with that workaround? I'm especially worried about
all the numerous variables in the SPM.mat file that this might break.
Thanks,
Ming Hsu
|