Dear Matthew,
thanks for the explicit answer, we have
done it the way you describe it. It is
really striking how a tiny variation
in the proportional scales dramatically
changes the results. Theoretically the
reason for this is very clear and has
been discussed very often, nevertheless
we were surprised.
We compared global proportional scaling
as done by spm with a proportional
scaling with the mean of the lateral
ventricles, an idea that has been
introduced by others.
The results are very much in favor of
the csf scaling.
Regards
Bruno
--
Bruno Weber
Neurologische Klinik &
Nuklear Medizin
UniversitaetsSpital
Frauenklinikstrasse 26
CH-8091 Zuerich
Tel: ++41 1 255 55 79
Fax: ++41 1 255 44 29
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|