Laminar MEG

MEG provides high temporal resolution, which is important for measuring rapid brain activity, but it has thought to suffer from worse spatial precision than fMRI. However, the spatial precision of MEG is theoretically only bounded by the precision at which the brain can be localized with respect to the MEG sensors. Typically, subjects move their head considerably during a scan and are not in the same position during each scan. This increases the accuracy of forward models used for source inversion, and limits the ability to combine data from multiple sessions. While a postdoc at UCL with Gareth Barnes & Sven Bestmann, Jimmy helped develop the use of subject-specific head-casts which drastically reduce within-session movement and between-session variability. We’re now developing techniques to use this high precision MEG data to:

  • Localize frequency-specific signals to deep or superficial cortical laminae
  • Estimate the orientation of cortical columns
  • Determine the time course of laminar activity in event-related responses
This project has received funding from the Fondation pour l'Audition (FPA)