Validating MEG and EEG finite element head models using a controlled rabbit experiment of skull defects

Abstract

Epilepsy affects 20 million people world-wide. When treatment of focal epilepsy with anti-epileptic drugs is ineffective, resective surgery may be considered. It is then essential to accurately determine the location of the seizure focus. Magnetoencephalography (MEG) and electroencephalography (EEG) allow us to reconstruct the location of event-related brain activity using a volume conductor model of the head. The objective of this thesis is to validate MEG and EEG finite element head models using a rabbit experiment of skull defects.

An in vivo rabbit experiment was developed that allowed recording high-resolution MEG and EEG above two conducting skull defects. An implantable, coaxial current source was constructed and placed at a series of positions in the cortex under the skull defects. An agarose gel was developed that provided a time-stable conductivity that mimicked different tissue types in the skull defects. A node-shifted, cubic finite element mesh of the head was generated, which differentiated nine tissue types.

For the first time, in vivo, experimental evidence was provided of the substantial influence of skull defects on MEG signals. The MEG signal amplitude reduced by as much as 20%, while the EEG signal amplitude increased 2-10 times. The MEG signal amplitude deviated more from the intact skull condition when the source was central under a skull defect. Using the exact finite element head model, forward simulation of the MEG and EEG signals replicated the experimentally observed characteristic magnitude and topography changes due to skull defects. When skull defects, with their physical conductivity, were incorporated in the head model, location and orientation errors during reconstruction were mostly eliminated. The conductivity of the skull defect material non-uniformly modulated its influence on MEG and EEG signals and source reconstruction.

The concordance of experimental measurements of the influence of skull defects on MEG and EEG signals and finite element simulations of exactly that experiment validated the finite element head modelling technique. Detailed finite element head models can improve non-invasive MEG- and EEG-based diagnostic localisation of brain activity, such as epileptic discharges.