[log in to unmask] wrote:
Hi Anna,
thanks for the quick answer!
> Hi Robert,
>
> You need to choose compare groups 1 scan per subject.
> Enter normal database (usually N>16) as group 1 and
> your individual patient scan as group 2.
If all things were so simple. Thanks.
> Its up to you whether you choose ANCOVA or proportional
> scaling for voxel intensity normalisation but I guess
> if the images are functionalised and therefore the
> global values are not too different between subjects
> then choose ANCOVA. Hope this helps.
I'm not sure what "functionalized" means. I normalize
and then smooth the images and they are all done on
the same scanner, so I suspect them to be similar.
One problem I have not solved is that age seems to
play an important role in glucose metabolism (as it
is also mentioned in the Signorini, Friston 1999
paper in NeuroImage) and I would like to know how
much of the effect can be attributed to that. The
paper says "we chose the right model" which is what
I want to do, as well :-)
My understanding is that I have to make age a
covariate (?) so that the age appears as a third
column in the design matrix (?). If that is the
correct way how do I proceed then?
Thanks for any hint again,
Robert
--
Robert Roebling, MD <[log in to unmask]>
|