Correspondence between regions of the DKT vs Harvard-Oxford atlas

Dear NeuroStars community,

I have a network of interest described in terms of cortical regions from the Harvard-Oxford atlas. I would need to describe this network in terms of parcellated regions that make up the surface-based DKT (Desikan–Killiany–Tourville) atlas, as implemented in FreeSurfer.

I have however not found (hints of) any correspondence table between the two atlases; I looked in the DKT repository (, its white paper (Klein&Tourville 2012) and the FS website. How could I establish a correspondence between regions, not 1:1 of course but quantifying the transformation’s accuracy (amount of overlap) for each parcelated region?

Assuming there is no quantitative correspondence/transformation, how would one go about at least eyeballing it, e.g. perhaps by doing a volumetric projection of the DKT and viewing that superimposed on the Harv.-Oxf., so I can see both parcellations at the same time?

Many thanks, and best wishes,