Non-linear parameter estimates from non-stationary MEG data

ABSTRACT: We demonstrate a method to estimate key electrophysiological parameters from resting state data. In this paper, we focus on the estimation of head-position parameters. The recovery of these parameters is especially challenging as they are non-linearly related to the measured field. In order to do this we use an empirical Bayesian scheme to estimate the cortical current distribution due to a range of laterally shifted head-models. We compare different methods of approaching this problem from the division of M/EEG data into stationary sections and performing separate source inversions, to explaining all of the M/EEG data with a single inversion. We demonstrate this through estimation of head position in both simulated and empirical resting state MEG data collected using a head-cast.

• Autores: López Hincapié, José David; Castellanos Domínguez, César Germán; Barnes, Gareth Robert; Baker, Adam; Woolrich, Mark W.
• Tipo de recurso: Article of investigation
• Fecha de publicación: 2016
• Institución: Universidad de Antioquia
• Repositorio: Repositorio UdeA
• Idioma: eng
• OAI Identifier: oai:bibliotecadigital.udea.edu.co:10495/7662
• Acceso en línea: http://hdl.handle.net/10495/7662
• Palabra clave: MEG inverse problem; Co-registration; Hidden Markov Model; Non-stationary brain activity; Bayesian comparison
• Rights: openAccess
• License: https://creativecommons.org/licenses/by/2.5/co/