A method for combining MEG and EEG to determine the sources.

Area of Science:

• Neuroscience
• Biophysics
• Medical Imaging

Background:

• Determining the location of neural activity (the inverse problem) is crucial for understanding brain function.
• Magnetoencephalography (MEG) and electroencephalography (EEG) are non-invasive techniques that measure brain activity.
• MEG is sensitive to tangential sources, while EEG detects both radial and tangential sources.

Purpose of the Study:

• To present and validate a novel three-step method for solving the neural inverse problem.
• To combine MEG and EEG data for improved source localization accuracy.
• To assess the method's robustness against noise and modeling errors using computer simulations.

Main Methods:

• A three-step computational method was developed to integrate MEG and EEG data.
• The method leverages the complementary sensitivity of MEG (tangential sources) and EEG (radial and tangential sources).
• Computer simulations were employed to test the method's performance with varying levels of noise and modeling uncertainties.

Main Results:

• The combined MEG-EEG method successfully localized a single dipole source in simulated data.
• The method demonstrated robustness with up to 5% root-mean-square (RMS) noise in MEG and EEG maps.
• Performance degraded significantly at 10% RMS noise, indicating sensitivity to higher noise levels.
• Limited tests with reasonable modeling errors (grid size, head radius, coil distance) showed the method still performed well.

Conclusions:

• The presented three-step method offers a viable approach for solving the neural inverse problem by integrating MEG and EEG.
• The technique is effective under realistic noise conditions (up to 5% RMS noise).
• Further investigation is warranted to fully understand the impact of significant noise and complex modeling errors on performance.