Dear Gordon,
I think all of these approaches would be valid, though grouping studies by
by gyri or Brodmann areas alone can some times be too general. Most gyri
and Brodmann areas are rather large, and often contain several (portions of)
functional subregions within them, so grouping activations in this way might
conceal interesting diferences. (And of course, there is always the risk
of having functional subregions that do obey neither gyri nor Brodmann
area borders, but that's a diferent issue altogether.)
But in my experience (mostly working on prefrontal cortex) I have found
that it's sometimes useful to consider the regions of intersection between
Brodmann areas and gyri (provided that the studies you're reviewing used
normalization to Talariach space and reported Talariach cordinates). For
instance, the intersection between middle frontal gyrus and BA46 may be one
region; the intesrection between BA47 and inferior prefrontal gyrus may be
another region, and so on.
Sometimes it also helps to know a bit about previously proposed functional
subdivisions - information you could use to reduce the number of possible
regions formed by intersections. For instance, the intersection between
BA46+BA9 and middle+superior frontal gyri would correspond to the dorsolateral
prefrontal cortex; the intersection between BA47+BA45 and inferior
prefrontal gyrus would correspond to the ventrolateral prefrontal cortex,
and so on.
Once you have decided on such regions or subregions, you could classify
the reported foci, according to regions and tasks.
Also, I've found it very useful to note down not only the target task
for which activations were observed, but also the baseline task. Very
often, differences in baseline tasks may produce differences in
activation, although the same target task was employed.
Kalina
On Fri, 3 Aug 2001, Gordon Waiter wrote:
> SPM'rs
>
> We are about to start a new study and want to compare the results from a
> number of previous studies looking at similar processes. My question is,
> at what resolution, such as gyrus, Brodmann area, or Talairach
> co-ordinate is it valid to do a meta-analysis, of activating regions in
> those published studies, ie group all regions across studies together.
>
> Thanks,
>
> Gordon
>
> --
> Dr. Gordon D. Waiter
> Cognitive Neuroimaging Research Unit
> Dept. of Psychology
> University of Aberdeen
> King's College
> Aberdeen
> AB24 2UB
>
> 01224 272234/559725
>
> [log in to unmask]
>
_____________________________________________________________________________
Kalina Christoff, Ph.D. Email: [log in to unmask]
Office: Rm.422; (650) 725-0797
Department of Psychology Home: (408) 245-2579
Jordan Hall, Main Quad Fax: (650) 725-5699
Stanford, CA 94305-2130 http://www-psych.stanford.edu/~kalina/
_____________________________________________________________________________
|