Conservative Finite Element Modeling of EEG and MEG on Unstructured Grids

Area of Science:

• Computational neuroscience
• Biomedical engineering
• Electrophysiology

Background:

• Accurate modeling of forward problems is crucial for interpreting electroencephalography (EEG) and magnetoencephalography (MEG) data.
• Existing methods may face limitations in accuracy, particularly with lower-resolution head models.

Purpose of the Study:

• To assess the performance and accuracy of the mixed-hybrid finite element method (MHFEM) for EEG and MEG forward modeling.
• To compare MHFEM with the conventional nodal finite element method (P1 FEM).

Main Methods:

• Applied MHFEM to EEG and MEG modeling using unstructured tetrahedral grids.
• Utilized a layered spherical head model and a realistic head model for simulations.
• Employed a subtraction approach to handle singular sources in magnetoencephalography (MEG) computations.

Main Results:

• MHFEM provides approximate potential and induced currents, ensuring discrete charge conservation.
• The method results in a positive semi-definite matrix, solvable by standard iterative techniques.
• MHFEM demonstrated higher accuracy than P1 FEM on low-resolution head models.

Conclusions:

• MHFEM is a conservative and accurate method for EEG and MEG modeling, especially on low-resolution models.
• Despite higher computational cost, MHFEM's accuracy justifies its use where P1 FEM falters.