Dear SPMers,
I know that anatomical and cytoarchitectural referencing is a persistent and often frustrating problem in functional neuroimaging, and appreciate very helpful postings from authorities on these issues. I am looking for any collective wisdom about the following issue. I am working with a colleague to make inferences about anatomical locations for an MNI-based template. Thus far, our search seems to reveal that the Lancaster transformation (Neuroimage. 2010 June; 51(2): 677–683) is the best available option for Brodmann's areas (though clearly there is great error due to individual variability and the nature of the transformation). As mentioned on these forums, probablistic atlases are considered preferable, but there are many available constructed for various purposes. Given that there are considerable drawbacks to Brodmann's identification, we'd appreciate greatly any feedback on the following:
1) Is the Lancaster transformation the best of currently available options for MNI to Talairach Brodmann's estimation?
2) In lieu of procedures that attempt to address Brodmann's locations via transformation, what is a concensus high-quality probabilistic reference for either Brodmann's regions or anatomically specific divisions (inferior, middle, frontal gyrus, etc) that could suffice for defining anatomical specificity in MNI space? Is WFU Pickatlas the most trusted way to go for whole brain atlasing in SPM?
3) Regardless of 1) or 2) above, does anyone know about or would be willing to share a script that is capable of extracting the MNI location of every voxel within a defined mask in space? That is, a script that could extract the location for every voxel in a mask based on task activation?
Thanks very much for any help,
John Medaglia
Pennsylvania State University
|