Head-Specific Spatial Spectra of Electroencephalography Explained: A Sphara and BEM Investigation.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Electroencephalography (EEG) is crucial for neuroscience research, but analysis relies on simplified head models.
• Accurate spatial-frequency analysis of EEG is essential for understanding brain activity and designing neuroimaging techniques.
• Previous methods were limited by the assumption of spherical head models.

Purpose of the Study:

• To develop and validate a novel method for spatial frequency analysis of EEG data using realistic head models.
• To compare the proposed method with existing techniques on spherical head models.
• To determine the impact of spatial sampling density on EEG power estimation.

Main Methods:

• Utilized the Sphara technique for spatial Fourier analysis on arbitrarily shaped surfaces.
• Employed a five-compartment Boundary Element Method (BEM) head model for realistic volume conductor simulation.
• Validated Sphara against discrete spherical harmonics on a spherical volume conductor.
• Derived signal-to-noise ratio (SNR) requirements for EEG spatial sampling.

Main Results:

• The Sphara method was validated and uncertainty limits were established.
• Conventional EEG electrode placement (e.g., 10-20 system) can lead to significant EEG power misestimation (up to 50%).
• Even with 64 electrodes, EEG power estimation errors can reach up to 15%.

Conclusions:

• The proposed Sphara method enables accurate spatial frequency analysis of EEG on realistic head geometries.
• Current EEG sampling strategies may introduce substantial errors in power spectrum estimation.
• Findings offer insights for optimizing EEG acquisition and transcranial electric stimulation targeting.