Zeynep Akalin Acar

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.

Accurate EEG source localization requires a realistic structure-based electrical forward model of the subject’s head. We performed a simulation study to investigate the effects of different head models, electrode co-registration errors, and skull conductivity modeling errors on EEG source localization, and asked which brain regions are more sensitive to the choice of head model modeling errors. A realistic Boundary Element Method (BEM) model generated from a subject's T1-weighted MR image was first used to generate simulated EEG scalp projections for a Cartesian grid of 6,717 cortical sources. The source locations were then estimated from these projection maps using three more generic head models generated using the Neuroelectromagnetic Forward head modeling Toolbox (NFT; Akalin Acar & Makeig, 2010): a spherical model, an MNI template model, and this model warped to the measured 3-D EEG sensor locations. The position and orientation of each of the source dipoles were then estimated within the other three head models. Though the MNI models better captured the overall head shape, because the MNI head image lacks mouth and jaw structures, location errors (up to 4 cm) were largest in inferior brain regions, particularly for z-oriented sources. Mean localization error was 14.3 mm for the electrode-warped MNI model, 14.9 mm for the standard MNI model, and 15.1 mm for the three-layer spherical model. To understand the effect of skull conductivity errors, we decreased the skull/brain conductivity ratio in the realistic head model from 1/20 to 1/80; this produced largest location errors (up to 2 cm) for dipoles near the skull.