A full subtraction approach for finite element method based source analysis using constrained Delaunay...

Area of Science:

• Biophysics
• Computational Neuroscience
• Medical Imaging

Background:

• The dipole model is standard for electroencephalography (EEG) source analysis.
• The governing Poisson-type differential equation presents a singularity challenge with dipole models.
• Accurate modeling of current sources is crucial for interpreting EEG data.

Purpose of the Study:

• To present a full subtraction approach for handling dipole singularities in EEG source analysis.
• To develop a method for fast and accurate inverse problem computation in anisotropic volume conductors.
• To improve the precision of EEG source localization.

Main Methods:

• A full subtraction approach dividing potential into singularity and correction potentials.
• Finite element (FE) method for computing correction potentials.
• Constrained Delaunay tetrahedralisation (CDT) for high-quality FE mesh generation.
• Careful evaluation of right-hand side integrals for high convergence order.

Main Results:

• Achieved maximal relative error of 0.71% in a complex four-layer sphere model.
• Validated the approach for radial and tangential sources with eccentricities up to 1 mm.
• Demonstrated high accuracy with a CDT-FE model of 360,000 nodes.
• The method is suitable for arbitrary dipole locations without local mesh refinement.

Conclusions:

• The full subtraction approach combined with CDT-FE meshes offers unprecedented accuracy in EEG source analysis.
• This method enables fast computation of the inverse problem for anisotropic head models.
• The findings advance the capabilities of EEG source localization techniques.