This is why I generally suggest doing the spatial normalization via the
Segment button. If you really want to use the Normalise button, then
maybe try segmenting the image first and spatially normalizing the GM
tissue class to a GM template (or possibly finding a skull stripped
template, and matching skull stripped versions of your images to this).
Best regards,
-John
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of Roman Rodionov
Sent: Thursday, June 21, 2007 3:40 PM
To: [log in to unmask]
Subject: [SPM] difference btw results of segmentation and normalisation
Dear John and List,
I think it is well known that results of segmentation and normalisation
operations for the same image produce different results. They are not
completely different but difference of about 3-5 mm is not satisfactory
in
our case.
My question is if there is a way (may be special set of options for
segmentation opreation) of producing the same result by these two
operations?
The reason we need it is getting _inv_sn.mat which would allow to go
from
normalised to individual space.
I found 2-3 questions similar to this (including our a year old one -
inversion of _sn.mat) but I didn't find an answer.
Thanks a lot,
Roman
|