Hi Tyler

The atlas is described in our 2002 (Neuron) and 2004 (NeuroImage) papers. It implements the anisitropic, nonstationary Markov Random Field segmentation. I think lots of people have compared their segmentations to it, although I would urge some caution as many of them are biased IMO (e.g. they use their own segmentation definitions or the MRI sequence they were trained on to drive the comparison).

Cheers

Bruce

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> On Behalf Of Rooks, Tyler
Sent: Monday, November 9, 2020 9:32 AM
To: freesurfer@nmr.mgh.harvard.edu
Subject: [Freesurfer] Subcortical Segmentation Atlas Question

        External Email - Use Caution        

Good Morning FreeSurfer Developers and Experts,

I hope all is well.  I am relatively new to the use of FreeSurfer and neuroimaging research in general.  I have been looking for any documentation or details regarding the probabilistic atlas used for subcortical segmentation tasks (RB_all_2008-03-26.gca) provided in the ‘average’ directory.  Is there any information available or that can be described about what this atlas was generated from or any additional details about the atlas?  I am looking for information relating to potential number of subjects used to develop the atlas if that is available or methods used to develop?  Additionally, are there any comparisons between this subcortical atlas and other subcortical atlases available?  I have found a few comparisons between FreeSurfer and other software/atlases, but nothing comparing the subcortical segmentation directly.

Any help or information would be greatly appreciated.

Thanks,

Tyler

Ph.D. Student 

Medical College of Wisconsin