Hi,

I refrained from answering until now but it seems the issue is not that 
interesting to the rest of the community as nobody else chimed in ;) 
Still, here's my 2 cents:

> In the following article AAl toolbox has been used for anatomic
> localization in a pediatric population

and

> Further in the following article, the MNI coordinates from a pediatric
> study were transformed into talairach coordinates using talairach daemon

seem to touch on 2 different issues. One, more technical, I believe it 
is obvious that the framework used to localize changes should be matched 
with the framework to normalize the data in the first place. I.e., if 
you used a custom template, using standard masks for an exact 
localization may be problematic. Of course, the overall delineation will 
likely be correct, but one may want to be wary regarding an exact 
localization. Along the same lines, I am always very hesitant to report 
coordinates when using a custom template as the assumption of a common 
reference frame is not valid anymore in my view (i.e., my [12 22 31] may 
mean something different as somebody else's same numbers as the 
shape/tissue composition of my reference data may differ from his, 
despite both being in the same standard space). Two, more biological, 
brain  development: if Brodman areas, for example, are already quite 
variable in adults, should we not expect them to be more variable still 
in children? The brain sure undergoes enormous changes during 
development, and expecting a one-to-one correspondence with an adult 
brain is certainly a vast oversimplification. Children are not small 
adults, and the same goes for the respective brains.

> I would appreciate comments on what is the best way of anatomical
> localization in pediatric images and whether the above provide a way of
> anatomical localization in pediatric sample .

I think there are rights and wrongs, but there is a lot of gray in 
between. It will depend on the exact implementation of a given approach 
that will decide whether it is more or less helpful in the context of a 
given study. Personally, I am simplistic by overlaying my results on my 
reference data (not on a single-subject brain) and using my neuroanatomy 
to describe where things are happening. This is of course a pain to 
those wanting to do meta-analyses, but it seems appropriate to me.

Hope this helps,
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
=====================================================================