Electrode and brain modeling in stereo-EEG.

OBJECTIVE: To quantify the perturbation due to the presence of a measuring depth electrode on the intracranial electric potential distribution, and to study the effect of the heterogeneity and anisotropy of the brain tissues' electric conductivity.
METHODS: The governing differential equations are solved with the Boundary Elements Method to compute the perturbation on the electric potential distribution caused by the presence of the measuring electrode, and with the Finite Elements Method to simulate measurements in an heterogeneous anisotropic brain model.
RESULTS: The perturbation on the measured electric potential is negligible if the source of electric activity is located more than approximately 1mm away from the electrode. The error induced by this perturbation in the estimation of the source position is below 1mm in all tested situations. The results hold for different sizes of the electrode's contacts. The effect of the brain's heterogeneity and anisotropy is more important. In a particular example simulated dipolar sources in the gray matter show localization differences of up to 5mm between homogeneous isotropic and heterogeneous anisotropic brain models.
CONCLUSIONS: It is not necessary to include detailed electrode models in order to solve the stereo-EEG (sEEG) forward and inverse problems. The heterogeneity and anisotropy of the brain electric conductivity should be modeled if possible. The effect of using an homogeneous isotropic brain model approximation should be studied in a case by case basis, since it depends on the electrode positions, the subject's electric conductivity map, and the source configuration. SIGNIFICANCE: This simulation study is helpful for interpreting the sEEG measurements, and for choosing appropriate electrode and brain models; a necessary first step in any attempt to solve the sEEG inverse problem.