Quantifying radial vs. tangential source composition per brain region

Support & Discussions
1

Hi all,

As you may know, MEG is predominantly sensitive to tangential sources. For my analysis, I need to correlate source-localised power across brain regions with a measure that captures the radial vs. tangential composition of each region. I have a few questions about the best approach:

  1. Would the cortical geometry statistics from FreeSurfer (e.g., lh.aparc.a2009s.stats) be a valid proxy for this? If so, which metric would be most appropriate?

  2. Alternatively, should I derive orientation information directly from my dipoles? I’m using a BEM forward model with a scalar beamformer ; what would be the correct way to extract and summarise dipole orientations per region in this setup

Thanks in advance!