Print

Print


Dear Sarah,

> 1) The creation of the templates. The templates are an average of the
> structural scans I input, correct?  For example: When looking for template 6
> to run normlisation, there is only one template 6, in the folder of the
> first of the 21 subjects I entered in the new segment volumes.  It looks to
> be an average of the scans.

Yes, that's right.


> 2) If the templates created are an average, should I preprocess ALL of my
> data in one batch?  I have one group of healthy adults, and another
> clinical/patient group.  I will be looking at group differences. I'm not
> 100% clear on what the template does, so I'm not clear on whether I need to
> process everyone together, or if I need to process each group separately?

The segmentation is independent for each subject, but you're right
that the template creation should be done with all of your data so
that everyone contributes to the template (with one small caveat, see
below).

The idea is that DARTEL creates an average brain from all of your
subjects.  The warp from each subject's native space to this template
space is computed as the template is created; at the end of template
creation, then, you have the template itself, but perhaps more
importantly you have the warps that take each subject into this common
space.  (These warps can be used to help bring subjects into MNI space
as well - which is just a different common space.)

In general, you want all of your subjects in your template.  The
exception might be if you have more than 1 group of subjects, as in
your case, and the groups are not equal in size. In this case I would
randomly select equal numbers from each group to create a template,
and then normalize everyone to this template.  The reason for doing
this is that you don't want your template to be biased towards one
group more than the other (which may in turn exert some bias on the
normalization).

Hope this helps!

Best regards,

Jonathan

-- 
Dr. Jonathan Peelle
Center for Cognitive Neuroscience and
Department of Neurology
University of Pennsylvania
3 West Gates
3400 Spruce Street
Philadelphia, PA 19104
USA
http://jonathanpeelle.net/