Global sensitivity of MEG source analysis to tissue conductivity uncertainties

Area of Science:

• Biophysics
• Neuroimaging
• Computational Neuroscience

Background:

• Inter-individual variations in tissue conductivity are crucial for electroencephalography (EEG) source analysis.
• Their influence on magnetoencephalography (MEG) source analysis is less understood and often assumed to be minimal.

Purpose of the Study:

• To conduct an in-depth analysis of the influence of tissue conductivity variations on MEG forward solutions and source analysis.
• To compare these effects with those observed in EEG source analysis.
• To improve the interpretation of MEG results and combined EEG-MEG analyses.

Main Methods:

• Simulated MEG forward solutions using a detailed five-compartment head model and FEM multipole approach.
• Employed a generalized polynomial chaos approach (gPC) to calculate MEG leadfields for varying tissue conductivities.
• Investigated the sensitivity of MEG forward solutions and source analysis to conductivity uncertainties.

Main Results:

• MEG is significantly more robust to tissue conductivity uncertainties than EEG.
• Conductivity variations notably affect the topography of MEG forward solutions for quasi-radial sources.
• Significant localization errors (up to 5 mm) and effects on source orientation/magnitude occur for tangential sources, particularly without rank reduction.

Conclusions:

• While MEG is generally robust, tissue conductivity uncertainties should not be entirely neglected in source analysis.
• Effects on quasi-tangential sources can have practical implications for combined EEG-MEG analyses.
• Careful consideration of conductivity variations is needed for accurate MEG source localization and interpretation.