Dear Kerlany,
On Wed, 10 Jun 2009 16:45:42 +0200, kerlany pereira <[log in to unmask]> wrote:
>Hi VBM users,
>I did two jobs, in VBM5 - 'Estimate and Write',with the same parameters, and
>the only difference is:
>For the first job I used 3 images and for the second job I used only 1
>image. For the first job the image 1 of 3 images submitted a terrible
>result, but the result for the same image in a second job is very good.
This is really strange because the segmentation should be independent for each subject.
What I have noticed in your screenshot is that for one run your T1-image is skull-
stripped, but for the other not. Because this makes no sense at all I would first try to
update spm5 and vbm5 to the newest version and try the segmentation again (standard
procedure if we are running out of ideas...).
>In 'Affine Regularisation' I put 'Average sized template' . Is there a
>problem when it calculates the average? OBS: the sequences of MRI has the
>same parameters and the patient's age is 10-14 years.
>And, what's the difference between 'No regularisation' and 'No affine
>registration'?
Even for children data I would use the default ICBM space template regularization.
Regularization is thought to make the spatial registration more robust by penalizing
stretching or shrinking. However, the starting estimates are often more important. If
registration fails it often helps to roughly define the origin (AP) in your images. If you use
VBM5 keep in mind to deselect the option "use center of mass to set origin" if you want to
manually define the origin.
Regards,
Christian
--
____________________________________________________________________________
Christian Gaser, Ph.D.
Assistant Professor of Computational Neuroscience
Department of Psychiatry
Friedrich-Schiller-University of Jena
Jahnstrasse 3, D-07743 Jena, Germany
Tel: ++49-3641-934752 Fax: ++49-3641-934755
e-mail: [log in to unmask]
http://dbm.neuro.uni-jena.de
>I'm sending the images.
>Thanks in advance,
>Kerlany Pereira
>
|