Ghost interactions in MEG/EEG source space: A note of caution on inter-areal coupling measures

Area of Science:

• Neuroscience
• Biophysics
• Signal Processing

Background:

• Magnetoencephalography (MEG) and electroencephalography (EEG) enable non-invasive study of long-range cortical interactions.
• Inter-areal connectivity estimation is challenged by field spread and volume conduction, causing artificial linear correlations.
• Standard connectivity measures are affected by linear source mixing, impairing source separability.

Purpose of the Study:

• To investigate the impact of field spread and volume conduction on advanced connectivity measures.
• To demonstrate the generation of spurious "ghost" interactions in MEG/EEG source space analyses.
• To quantify the effect of signal mixing on connectivity measures, including those insensitive to zero-lag correlations.

Main Methods:

• Simulations were used to quantify the effects of field spread and volume conduction.
• Analysis focused on phase- and amplitude-correlation based connectivity measures, such as imaginary coherence and orthogonalized amplitude correlation.
• The separability of neuronal phase and amplitude correlations under signal mixing was assessed.

Main Results:

• Advanced connectivity measures, while immune to direct linear mixing, can still produce significant numbers of false positive connections.
• Spurious "ghost" interactions arise from field spread near true interactions, affecting both region-of-interest and connectome analyses.
• Signal mixing fundamentally limits the separability of neuronal phase and amplitude correlations.

Conclusions:

• Spurious correlations pose a significant challenge in MEG/EEG source space connectivity analyses.
• Even connectivity measures designed to overcome linear mixing effects are susceptible to false positives.
• Careful consideration of "ghost" interactions is crucial for accurate interpretation of MEG/EEG connectivity findings.