Data-driven Fokker-Planck models of macroscopic cortical activity as measured with MEG

Rikkert Hindriks

VU Amsterdam

Cortical rhythms as measured with EEG/MEG reflect the synchronous post-synaptic activity of a large number of pyramidal cells. They are a collective phenomena, emerging from the complex interactions between large numbers of neurons. As such, their intrinsic as well as their synchronization dynamics might be described by Fokker-Planck equations. Although Fokker-Planck equations are successfully fitted to recordings from various complex natural systems, their applications to neurophysiological recordings are currently very limited. Moreover, the methodology needed to obtain statistically sounds results is commonly not exploited in practice. We describe how Fokker-Planck equations can be fit to EEG/MEG data and discuss the methodological issues that arise. We apply the methods to assess learning-related changes in the synchronization between primary motor cortices. The results provide insight into the deterministic and stochastic dynamical principles underlying the coordination of cortical rhythms.

Back