Hi Julien,

the RB atlases are derived from 40 manually labeled subjects provided to use by Randy Buckner (hence the "RB"). I think we now user a somewhat newer once, but in principle it is similar to the RB_all_2006-2-15 atlas. We also have probabilistic surface-based architectonic maps for a subset of the areas done by the Juelich/Dusseldorf groups that we distribute (see the recent Cerebral Cortex paper on our website).

cheers, Bruce

On Fri, 12 Dec 2008, Julien Dubois wrote:

Hi freesurfer developers

I'm working with freesurfer's subcortical anatomical labels. I understand that they are based on a database of manually labeled anatomical scans, which were normalized to the MNI template in order to get probability maps, that freesurfer uses as prior information. In a slide (in freesurfer.recon.2007.ppt I believe) there is a line that says the atlas is RB_all_2006-2-15. What is this particular atlas based on? how many brains were manually labeled? I couldn't find any reference for this atlas.

There exist some probabilistic cytoarchitectonic maps (cf Eickhoff NeuroImage 2005). I understand that freesurfer only makes use of manually labeled MRI scans, not cytoarchitectonic information. Is that right? how hard would it be to use the probabilistic cytoarchitectonic atlas instead of RB_all_2006-2-15? is it doable at the user level or does it require messing around with the code?

Thank you very much in advance!

• Julien

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer