Geometry-adapted hexahedral meshes improve accuracy of finite-element-method-based EEG source analysis

Area of Science:

• Biomedical Engineering
• Computational Neuroscience
• Medical Imaging

Background:

• Finite-element (FE) method-based electroencephalography (EEG) source analysis accuracy is highly dependent on mesh generation.
• Optimizing meshing strategies is crucial for balancing accuracy in potential distributions with computational efficiency.

Purpose of the Study:

• To improve the accuracy of EEG source analysis by proposing a novel mesh generation strategy.
• To evaluate the effectiveness of smoothing regular hexahedral finite elements at material interfaces using a node-shift approach.

Main Methods:

• Presented theoretical frameworks for modeling current dipoles in FE volume conductors using subtraction and direct potential methods.
• Evaluated regular and smoothed elements in a four-layer sphere model for both potential approaches.
• Computed and visualized potential distributions using regular and geometry-adapted hexahedra FE models of the human head.

Main Results:

• Node-shifting reduced topography and magnitude errors by over a factor of 2 for tangential and 1.5 for radial sources.
• The node-shift approach improved accuracy for both subtraction and direct potential methods.
• Caution is advised for node-shifting near irregular hexahedra, especially with the subtraction method, to avoid increased errors.

Conclusions:

• Node-shifted hexahedra are recommended for skin and skull compartments in realistic head models.
• Smoothing finite elements via node-shifting significantly enhances EEG source analysis accuracy.
• Deforming elements at grey and white matter surfaces is not recommended due to potential error increases.