Source to sensor coupling (SoSeC) as an effective tool to localize interacting sources from EEG and MEG data.

Area of Science:

• Neuroscience
• Biophysics
• Signal Processing

Background:

• Conventional methods for estimating interacting brain sources from EEG/MEG data involve calculating pairwise voxel coupling, which is computationally intensive.
• These standard approaches can be prohibitively slow, especially when bias removal is required.

Purpose of the Study:

• To introduce a novel, computationally efficient method for identifying interacting neural sources from electroencephalography (EEG) and magnetoencephalography (MEG) data.
• To address the computational cost and potential biases associated with traditional source interaction estimation techniques.

Main Methods:

• Proposed a new approach that replaces pairwise voxel coupling with maximization of coupling between each source and the sensor space signal.
• Utilized the imaginary part of coherency as the coupling measure to avoid self-coupling confusions.
• Discussed conceptual aspects related to vector beamformers and eLoreta.

Main Results:

• The proposed method robustly detects coupled sources in both simulated and empirical EEG data.
• Demonstrated that the new approach is hundreds of times faster than conventional methods.
• EEG resting-state data analysis revealed greater statistical power compared to existing approaches.

Conclusions:

• The novel source-to-sensor coupling maximization method is an effective tool for identifying interacting sources from EEG and MEG cross-spectra.
• This technique offers significant improvements in speed and statistical power for neural connectivity analysis.