Effects of Forward Model Errors on EEG Source Localization

Abstract

Subject-specific four-layer boundary element method (BEM) electrical forward head models for four participants, generated from magnetic resonance (MR) head images using NFT (www.sccn.ucsd.edu/wiki/NFT), were used to simulate electroencephalographic (EEG) scalp potentials at 256 recorded electrode positions produced by single current dipoles of a 3-D grid in brain space. Locations of these dipoles were then estimated using gradient descent within five template head models fit to the electrode positions. These were: a spherical model, three-layer and four-layer BEM head models based on the Montreal Neurological Institute (MNI) template head image, and these BEM models warped to the recorded electrode positions. Smallest localization errors (4.1–6.2 mm, medians) were obtained using the electrode-position warped four-layer BEM models, with largest localization errors (~20 mm) for most basal brain locations. When we increased the brain-to-skull conductivity ratio assumed in the template model scalp projections from the simulated value (25:1) to a higher value (80:1) used in earlier studies, the estimated dipole locations moved outwards (12.4 mm, median). We also investigated the effects of errors in co-registering the electrode positions, of reducing electrode counts, and of adding a fifth, isotropic white matter layer to one individual head model. Results show that when individual subject MR head images are not available to construct subject-specific head models, accurate EEG source localization should employ a four- or five-layer BEM template head model incorporating an accurate skull conductivity estimate and warped to 64 or more accurately 3-D measured and co-registered electrode positions.

EEG Source Reconstruction in Male Nonsmokers after Nicotine Administration during the Resting State

“…The forth limitation regards the fact that the previous EEG literature [41] suggested that a four-layer BEM of the head wrapped on a single subject's MRI scan should provide the minimal localization error (mean 5.4-mm error). Instead, in the present study, a three-layer BEM of the head was wrapped on a MNI template scan for all subjects.…”

“…The statistical inference chosen was a cluster inference analysis on pseudo t statistic values [36]. Although cluster inference analysis suffers from low spatial specificity when significant clusters are large [40], this is not considered an issue for the already low spatial resolution of EEG source reconstruction methods [41]. The smoothing at statistical level was again set at 32 mm in order to comply with the smoothing already used for computing the sources.…”

Mathematical Foundation of Electroencephalography

“…Whereas boundary element models are adequate to portray major tissue compartments, such as the cerebrum and skull, they fail to represent detailed anatomical information within the compartments, such as the cerebral folding [14,15]. Finite element models, on the other hand, are efficient in capturing these details, but are labour intensive and computationally demanding [16,17].…”

Influence of the head model on EEG and MEG source connectivity analyses

“…However, to the best of our knowledge, none of these studies systematically considered the influence of the head model on source connectivity analysis, in spite of the fact that forward modeling errors are known to have a significant effect on the accuracy of source analysis (Acar and Makeig, 2013;Aydin et al, 2014;Dannhauer et al, 2011;Fuchs et al, 2007;Hallez et al, 2008;Haueisen et al, 1997Haueisen et al, , 2002Lanfer et al, 2012a,b;Montes-Restrepo et al, 2014;Stenroos et al, 2014;Wolters et al, 2006). Spherical head models (de Munck and Peters, 1993) or the boundary element method (BEM) using 3 layers for the skin, skull, and brain (Fuchs et al, 2007;Kybic et al, 2005) were used in most previous studies on source connectivity as a compromise between computational cost and accuracy.…”

“…Recently, several such techniques have been proposed, including BEM with 4 or 5 tissue layers (Acar and Makeig, 2013), the finite difference method (Hallez et al, 2005;Vanrumste et al, 2000), and the finite element method (FEM) (Buchner et al, 1997;Haueisen et al, 1997;Marin et al, 1998;Schimpf et al, 2002;van den Broek et al, 1998;Wolters et al, 2004). Among these, the FEM is the most versatile, as it allows the modeling of arbitrary anisotropic conductivity profiles using any type of discretization.…”