Hi Sophie
> Thanks John, I am ccing this to the list for help.
> I'm doing VBM on SPM5 right now.
> I have one group of scans from studies a couple of years ago
> (patients & controls) that the FOV was 30x30
> I have another group of scans from a study last year (also patients &
> controls) that the FOV was 24x24
I had kind of similar data
>
> right now I'm assuming that I can't put them all in the same VBM
> analysis since the voxel sizes are different?
> or could I put them all together and then add some sort of regressor
> to account for this?
> So John suggested I use a covariate of 0's and 1's to account for
> voxel differences, I tried this and it didn't work.
> I'm using 2-nd level 2 sample t-test and it's not letting me enter the
> different voxel sized scans in 1 group. any ideas?
> So now I'm wondering, is this possible? if so, what stat analysis do I
> use.
well obviously you have to segment and normaluze so at the end of the
day all volumes should have the same size and voxel size if you used the
same template - now what you should do is what John told you (and me
and probably others .. thanks to him): enter an extraregressor with 1
and 0 for the different original volume fov (and it does work).
Note you can look at this regressor for fun, and you'll see why you
should do that.
Best
Cyril
--
Pernet Cyril, PhD
Center for Cognitive Neuroimaging
Voice Neurocognition Laboratory
Department of Psychology
58 Hillhead Street
University of Glasgow
Glasgow G12 8QB
Scotland
[log in to unmask]
http://vnl.psy.gla.ac.uk/
tel: +44 (0)141 330 4965
|