Hi Jenny,
as it concerns kids, I somehow have the urge to answer ;)
> 1. When doing the “optimized” procedure in SPM99, we understood there
> were rationales for making study-specific whole head and a priori
> templates, such as intensity variation between MRI machines and
> differences between our patients and the subjects in the templates
> (child vs. adult, patient vs. healthy control). Our understanding
> is that with the new Segment, it is not necessary to make a
> study-specific whole head template because the normalization is
> based on the a priori images. Are the rationales for creating
> study-specific a priori templates no longer relevant?
Basically, the need for creating custom templates becomes more pressing
with the expected differences between the group that makes up the
templates/priors (healthy young caucasian adults) and the group under
investigation. And yes, no need anymore to create a whole head template.
> 2. John suggests “hundreds of subjects” would be proper for making
> custom priors. Is 160 appropriately large?
I have created pediatric templates with 148 and then 200 subjects, and
found these numbers rather convincing.
> 3. Has anyone tried it both ways (with SPM vs. custom a priori) and
> found one way was better than the other?
We looked at both effects (standard vs. optimized and custom vs.
standard priors) in Keller et al, NeuroImage 2004. In this study on
adults, the effect of a customized prior was minimal. However, as you
are dealing with children, the influence even under spm5 must be
expected to be stronger, and I would suggest creating a customized
template from all participants.
> If anyone can offer advice or share their experience with this, we would
> greatly appreciate it. The Segment procedure is taking a long time for
> each subject, and making custom priors would take twice as long, but we
> would still like to do it if there are reasons to prefer it.
I would go for it. How exactly one should do this is another intersting
question, though (too long to elaborate on here ;)
Best,
Marko
--
=====================================================================
Marko Wilke (Dr.med./M.D.)
[log in to unmask]
Universitäts-Kinderklinik University Children's Hospital
Abt. III (Neuropädiatrie) Dept. III (Pediatric neurology)
Hoppe-Seyler-Str. 1, D - 72076 Tübingen
Tel.: (+49) 07071 29-83416 Fax: (+49) 07071 29-5473
=====================================================================
|