A hybrid boundary element-finite element approach for solving the EEG forward problem in brain modeling

Area of Science:

• Computational Neuroscience
• Biomedical Engineering
• Neuroimaging

Background:

• The electroencephalography (EEG) forward problem is crucial for interpreting brain activity.
• Existing methods like Boundary Element Method (BEM) and Finite Element Method (FEM) have limitations in modeling complex head geometries and tissue properties.
• Accurate modeling is essential for advancing neuroimaging techniques.

Purpose of the Study:

• To introduce and validate a novel hybrid Boundary Element-Finite Element (BE-FE) method for solving the EEG forward problem.
• To leverage the strengths of both BEM and FEM for improved accuracy and efficiency.
• To enhance the understanding of brain function through advanced neuroimaging.

Main Methods:

• Developed a hybrid BE-FE approach, dividing the head model into homogeneous BEM regions (including sources) and heterogeneous anisotropic FEM regions.
• BEM was used for the brain and dipole sources, while FEM modeled other head layers.
• Validated the method using spherical and realistic MRI-based head models with varying tissue properties (isotropic/anisotropic) and dipole configurations.

Main Results:

• The hybrid BE-FE method demonstrated significant improvements in accuracy over FEM across various head models and dipole parameters.
• In spherical models, enhancements of at least 1.05% (RDM) and 38.31% (MAG) were observed compared to FEM.
• For anisotropic realistic head models, improvements reached up to 55.4% (RDM) and 89.3% (MAG) over FEM.

Conclusions:

• The hybrid BE-FE method offers a promising and accurate approach for solving realistic EEG forward problems.
• This method enhances neuroimaging capabilities by providing more precise modeling of head tissues and brain activity.
• The findings contribute to a better understanding of brain function through improved EEG data interpretation.