[MINC-users] global t-test

[Sylvain MILOT]

Hi again Anil,

concerning fmristat's multistat script, I'm not sure this will be feasible after all
since you want to look at a group effect, hence it won't be possible to include fixed subject
effect variables in this case since the independent GROUP variable will be a linear combination of the
SUBJECT effect variables. My impression is that the SD regularization process requires that
the fixed SUBJECT effect variables be included in the model but I havent yet played with this feature so don't
consider my answer to be definitive. The same reasoning applies to glim_image concerning the linear
dependency.

Sylvain

On Tue, 6 Dec 2005, Sylvain MILOT wrote:

>
> Hello Anil,
>
> I haven't dealt with this kind of data personnally but you may want to have a look
> here:
>
> http://wiki.bic.mni.mcgill.ca/index.php/FullTextSearch?s=vbm
>
> I don't know how up-to-date this is but I'm pretty certain you should find what you're
> looking for!
>
> Now concerning glim_image, it should be feasible to use it to do exactly what you want.
> I doubt you should use the pooled sd measure since it requires that the sd image be
> stationary. Using the voxel sd measure would be more appropriate.
>
> It should also be possible to use Keith Worsley's latest method via fmristat's
> multistat script which regularizes the random effects variance by smoothing the ratio of
> fixed+random/fixed effects SD volumes and multiplying back by the fixed effects SD. Keith calls
> this a form of local sd pooling. Look here for references http://www.math.mcgill.ca/keith/.
> If it can be tailored to PET data, it should be feasible to taylor it to VBM data. I'm pretty sure
> Jason Lerch has already done this and maybe he'll answer your call when he
> gets a chance. In the meantime look at the wiki!
>
> Hope this helps a little,
>
> Sylvain
>
>
>
> On Tue, 6 Dec 2005, Anil Ashok Kuchinad, Mr wrote:
>
> > Hello,
> >
> >
> >
> > I am running a VBM analysis and I wanted to obtain on statistic that I
> > am not sure how to find. I was curious if someone knew what command
> > would allow me to find an overall t-value for the comparison between all
> > grey matter voxel density in one group vs. another. I basically want to
> > find the overall difference in grey matter density across the whole
> > brain between two groups.
> >
> >
> >
> > I am aware that glim_image has a pooled function but I couldn't get it
> > to work for me. I am truly shooting in the dark whether I am to use
> > glim_image and then run something similar to find_peaks or
> > tstat_threshold later on.
> >
> >
> >
> > Hopefully this is possible!
> >
> >
> >
> > Anil
> >
> >
> >
> >
> >
> > ----------------------------------------------------
> >
> > Anil Kuchinad
> >
> > Centre for Research on Pain
> >
> > McGill University
> >
> > Strathcona Anatomy and Dentistry Building
> >
> > 3640 University St., Room M/19
> >
> > Montreal, Quebec , Canada
> >
> > H3A 2B2
> >
> > Tel: (514) 398-1271
> >
> > Email: anil.kuchinad at mcgill.ca
> >
> >
> >
> > _______________________________________________
> > MINC-users at bic.mni.mcgill.ca
> > http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
> >
>
> ---
> Sylvain Milot (sylvain at bic.mni.mcgill.ca)
>               (trinity at bic.mni.mcgill.ca)
> Brain Imaging Centre
> Montreal Neurological Institute
> Webster 2B, Room 208
> Montreal, Qc., Canada, H3A 2B4
> Phone  : (514) 398-4965, Fax: 398-8948
> Mobile : (514) 712-1768
> Office : 527 Av Des Pins O., Room 204
>          Montreal, Qc., H2W 1S4
>
>
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>

---
Sylvain Milot (sylvain at bic.mni.mcgill.ca)
              (trinity at bic.mni.mcgill.ca)
Brain Imaging Centre
Montreal Neurological Institute
Webster 2B, Room 208
Montreal, Qc., Canada, H3A 2B4
Phone  : (514) 398-4965, Fax: 398-8948
Mobile : (514) 712-1768
Office : 527 Av Des Pins O., Room 204
         Montreal, Qc., H2W 1S4