Quantifying Cerebellar Signal Detectability in MEG and EEG in Epilepsy Using Anatomically Informed Source Modeling

Area of Science:

• Neuroscience
• Biophysics
• Medical Imaging

Background:

• The cerebellum's role in epilepsy is recognized, but its electrophysiological signals are difficult to detect noninvasively using conventional magnetoencephalography (MEG) and electroencephalography (EEG).
• Challenges include the cerebellum's deep location, complex structure, and unfavorable source orientations, limiting signal-to-noise ratio (SNR) in standard recordings.

Purpose of the Study:

• To quantitatively assess cerebellar signal detectability across different neuroimaging modalities and sensor configurations.
• To investigate the impact of sensor placement, source orientation, and head size on cerebellar electrophysiological signal detection.

Main Methods:

• Analysis of clinical MEG and EEG data from epilepsy patients.
• Construction of subject-specific anatomical models for consistent forward modeling.
• Estimation of SNR using anatomically informed source modeling, comparing standard SQUID-MEG/EEG with simulated on-scalp optically pumped magnetometer (OPM) configurations.

Main Results:

• Clinical MEG and EEG showed lower cerebellar SNR compared to superficial cortical areas.
• Placing OPMs at SQUID-equivalent locations did not improve cerebellar SNR, highlighting depth as a key factor.
• Cerebellar-optimized OPM layouts significantly increased SNR in posterior cerebellar regions, especially in individuals with smaller head sizes.

Conclusions:

• Cerebellar signal detectability is primarily determined by anatomical depth and geometry, not just sensor proximity.
• Flexible, region-specific sensor layouts, like optimized OPMs, can substantially improve cerebellar SNR.
• This framework offers a principled approach for evaluating MEG/EEG sensitivity in deep brain structures, advancing human brain mapping.