> Now I'm running a VBM analysis(using Brain MRI)of schizophrenia and normal
> controls to compare grey and white matter volume respectively.
> When I come to the basic models (chosing two sample t-test) stage, I select
> the following
>
> no grand mean scaling
> threshold masking absolute 0.05
> implicit mask (ignore zero's) YES
> explicity mask images NO
> Global calculation OMIT
>
> Are these options correct?
I don't see anything obviously wrong with these options.
>
> and then, when I compare schizoprenic's white matter vol. with controls',
> above options are right, too ?
Remember that for white-matter, it is probably advisable to smooth much more
heavily. I believe that for conditions such as schizoprenia, the overall
shape of the brain is different, so it may be worth trying some sort of
multi-variate approach, rather than a mass-univariate one. I'm pretty sure
that for these brains, your choice of global correction, and whether or not
you "modulate" the data, will influence your results. It is therefore worth
thinking about exactly what question you want to ask of the data.
Best regards,
-John
|