Print

Print
Dear Marko,

You're right; coregistration had failed.  i checked registration for my subjects, but not for both halves of the experiment, which evidently I should have. 

>We used to do this manually by centering and orienting the brain along the anterior-posterior commissure line, and you can still do that (depending on how many subject you have).

I actually did this manually for all my subjects, although I think I made a mistake for this one -- the anatomical image is off center.  I have to look back at the original image and figure out where the problem arose.  

>perhaps also within session (particularly if subjects moved around in between acquisitions).

I have coregistered all my subjects, but I didn't consider the possibility of doing it within session.  Within session is perhaps overkill, but there would be a pretty strong argument for within "half" -- it was between the first and second half of the experiment that the subjects got out and moved around.  

Thanks for your help! I know how to proceed now.

Best,
Katie

On Mon, Sep 26, 2016 at 2:27 AM, Marko Wilke <[log in to unmask]> wrote:
Hello Katie,

this is so drastic a failure that it points to the initial affine registration having failed. To fix this, you should make sure the subjects brains are in rough alignment with the template, and with each other.

We used to do this manually by centering and orienting the brain along the anterior-posterior commissure line, and you can still do that (depending on how many subject you have). If you have a lot, I would rather suggest you try using a combination of coregistration, and potentially re-setting the origin to the center of mass (there is a function in the vbm8 toolbox, if I remember correctly).

Also, make sure you always coregister anatomical and functional images from each subject, perhaps also within session (particularly if subjects moved around in between acquisitions).

Hope this helps
Cheers
Marko


Katie Surrence wrote:

Dear SPM experts,

I was missing half the brain at the second level, and I've tracked the
problem down to a subject for whom DARTEL normalization has apparently
failed, though only for the second half of the experiment.
  (Participants got out of the scanner and walked around between halves,
so some settings may have changed.)  It fails by squashing the brain
down in the bottom half of the field of view.

My question is, okay, I know what the problem is, now what?  Try another
kind of normalization and see if it does better, report that I did so in
any resulting paper, and give that participant its own "different
normalization" regressor at the second level?  Exclude data for the
problematic half of the participant?  Or is there a way to fix this?

I am attaching two pictures, one of the unsmoothed normalization failure
and one of the smoothed.  The brains are centered in the field of view
until the normalization step.

Thanks for any help you can provide!

Best,
Katie

--
____________________________________________________
Prof. Dr. med. Marko Wilke
 Facharzt für Kinder- und Jugendmedizin
 Leiter, Experimentelle Pädiatrische Neurobildgebung
 Universitäts-Kinderklinik
 Abt. III (Neuropädiatrie)

Marko Wilke, MD, PhD
 Pediatrician
 Head, Experimental Pediatric Neuroimaging
 University Children's Hospital
 Dept. III (Pediatric Neurology)

Hoppe-Seyler-Str. 1
 D - 72076 Tübingen, Germany
 Tel. +49 7071 29-83416
 Fax  +49 7071 29-5473
 [log in to unmask]tuebingen.de

 http://www.medizin.uni-tuebingen.de/kinder/epn/
____________________________________________________