Influence of isotropic skull models on EEG source localization

Abstract

We investigated the influence of using simplified skull models on electroencephalogram (EEG) source localization. The simplified skull models were derived from CT and MR images, with isotropic conductivity modeled as either heterogeneous or homogeneous. A total of four simplified head models were compared against a reference model with a skull accurately segmented with CT images. Our results show that the use of a simplified geometry for the skull, can lead to errors of approximately 1 cm for sources located in the central and temporal regions of the brain.

Effects of uncertainty in head tissue conductivity and complexity on EEG forward modeling in neonates

“…The effect of forward model inaccuracies concerning the geometry and conductivity of the various compartments on the precision of EEG source localization methods have been widely investigated in adults [Akalin Acar and Makeig, 2013]. The main sources of inaccuracies in EEG source analysis are exclusion of cerebrospinal fluid (CSF) [Rice et al, 2013;Vorwerk et al, 2014], skull inhomogeneity [Dannhauer et al, 2011;Lanfer et al, 2012;Li et al, 2007;Marin et al, 1998;Montes-Restrepo et al, 2013;Vorwerk et al, 2014], inhomogeneity of brain tissues [G€ ullmar et al, 2010;Vorwerk et al, 2014], and uncertainty of head tissue conductivity values [Ramon et al, 2004]. In newborn infants, however, few studies have investigated the effect of these deficiencies on the accuracy of EEG forward modeling and source localization [Gargiulo et al, 2015;Lew et al, 2013;Roche-Labarbe et al, 2008] mainly due to the difficulties in creating realistic head models in neonates.…”