Dear Alphonso and other CONN users,

Hope all is well.

Sorry a bit of an odd question. I had a chat with a colleague who mentioned that the Harvard-Oxford atlas in CONN may be of too low resolution for whole-brain NBS connectivity analyses. 

As the H-O atlas was built from morphological data, with limited number of ROIs,  we were wondering how it actually maps to functional networks?

During our discussion, my colleague suggested to use a more functionally defined atlas (eg Schaeffer200).

Do you have any recommendations?

NBS results using H-O atlas would still be valid, even if the ROIs are  structurally defined and less spatially accurate, right? (although representing a clear limitation)

Thanks for your time. Looking forward to reading your opinion on this!

Yann