Validating MEG estimated resting-state connectome with intracranial EEG

Area of Science:

• Neuroscience
• Biophysics
• Medical Imaging

Background:

• Magnetoencephalography (MEG) is a key tool for mapping resting-state brain connectivity.
• MEG's ill-posed source imaging and limited spatial resolution lead to source leakage, complicating connectivity interpretation.
• Validating MEG connectivity against a more spatially precise method is crucial.

Purpose of the Study:

• To validate Magnetoencephalography (MEG)-derived brain connectivity using a normative intracranial EEG (iEEG) atlas.
• To assess the impact of different connectivity metrics, including those correcting for zero-lag effects, on MEG-iEEG correlation.
• To evaluate the reliability of MEG for studying resting-state brain networks.

Main Methods:

• MEG data from 45 healthy participants were analyzed.
• The MEG inverse problem was solved using the wavelet-maximum entropy on the mean method.
• Four connectivity metrics (AEC, OAEC, PLV, wPLI) were computed and compared spatially with an iEEG atlas across canonical frequency bands.

Main Results:

• Moderate spatial correlations were found between MEG and iEEG for Amplitude Envelope Correlation (AEC) and Phase Locking Value (PLV).
• Spatial correlations decreased significantly when using metrics that correct for zero-lag connectivity (Orthogonalized AEC and weighted-phase lag index).
• MEG exhibited higher zero-lag connectivity compared to iEEG, suggesting source leakage influences connectivity estimates.

Conclusions:

• While MEG can recover relevant brain connectivity patterns, moderate-to-low correlations with iEEG necessitate cautious interpretation of MEG connectivity results.
• Zero-lag correction methods reduce correlations, indicating a trade-off between mitigating source leakage and potentially removing true connectivity.
• The findings highlight the importance of considering the impact of source leakage and chosen metrics on MEG-based connectivity studies.