> Up until now, I have been using the default analysis threshold of 0.8
> for VBM analyses. I have recently run an analysis threshold of 0.05, and
> as expected, obtained more widespread abnormalities in a patient
> population.
>
> My question is simply which threshold do I use and why?
Take a look at the mask.img that is produced. This should cover the region
that you are interested in analysing. If you are not doing any "grand mean
scaling", or using "proportioanl scaling" for your global normalisation, then
I would suggest an absolute threshold of about 0.05.
If you include very low intensity voxels, then you may experience
problems because of non-normality of the data (more severe deviations
are found in these regions because the data can not be less than 0,
which is not the behavious of a normal distribution).
Best regards,
-John
--
Dr John Ashburner.
Functional Imaging Lab., 12 Queen Square, London WC1N 3BG, UK.
tel: +44 (0)20 78337491 or +44 (0)20 78373611 x4381
fax: +44 (0)20 78131420 http://www.fil.ion.ucl.ac.uk/~john
|