Abstract
Event Abstract Back to Event Fokker-Planck dynamics of macroscopic cortical activity as measured with MEG Rikkert Hindriks1*, Fetsje Bijma1 and Aad V. Vaart1 1 VU University Amsterdam, Netherlands Macroscopic electrical activity emerging from human cortex is known to be involved in several cognitive and behavioral processes. Currently there is no consensus however, on its precise role. Understanding the dynamical principles underlying the macroscopic patterning of brain activity is an important aspect of research on the neuronal underpinnings of cognition and behavior. In this study we investigate the use of Fokker-Planck equations to capture the dynamics of EEG/MEG recordings. A Fokker-Planck equation describes the time-evolution of the probability density of the solution to a stochastic differential equation. It contains a deterministic term (drift) and a stochastic term (diffusion) and the resulting dynamics is a result of the interplay between drift and diffusion. Although Fokker-Planck equations have been fitted to time-series of several complex systems, the statistical issues are not always dealt with in a correct way. Moreover, their application to neuroscientific datasets is limited to a handful of studies [1]. In this study [2] we discuss three important issues that arise when estimating drift and diffusion parameters from a given time-series and propose methodology to deal with these issues. Moreover, we apply the proposed methodology to several MEG experiments. First, the construction of consistent estimators provides a challenge because we observe the continuous-time process only at discrete time points. This finite-time effect becomes an issue especially when the sampling frequency is low. We adopt the method proposed in [3] which yields consistent estimators irrespective of the sampling frequency. Secondly, we propose a strategy for model verification by â¨using appropriate residuals. And finally, the asymptotic â¨theory in [3] is adapted to provide confidence intervals for â¨the estimated model parameters.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
