Related Experiment Video
Updated: Apr 15, 2026
Network Analysis of the Default Mode Network Using Functional Connectivity MRI in Temporal Lobe Epilepsy
Published on: August 5, 2014
A symmetric multivariate leakage correction for MEG connectomes
G L Colclough1, M J Brookes2, S M Smith3
1Oxford Centre for Human Brain Activity (OHBA), University of Oxford, Oxford, UK; University of Oxford, Dept. Engineering Sciences, Parks Rd., Oxford, UK; Centre for the Functional Magnetic Resonance Imaging of the Brain (FMRIB), University of Oxford, Oxford, UK.
This study introduces a novel symmetric orthogonalisation method to eliminate spurious correlations in magnetoencephalographic (MEG) data. This technique accurately reconstructs functional brain networks from corrected brain regions of interest (ROIs).
Area of Science:
- Neuroscience
- Biophysics
- Computational Biology
Background:
- Magnetoencephalography (MEG) source reconstruction can introduce artificial correlations between brain regions.
- These spurious correlations complicate the accurate inference of functional brain networks (connectomes).
Purpose of the Study:
- To develop and validate a symmetric orthogonalisation method to correct for spurious correlations in MEG source data.
- To enable robust network modeling and connectome inference from corrected regions of interest (ROIs).
Main Methods:
- Proposed a symmetric orthogonalisation technique to remove artificial correlations among estimated source time-courses.
- Applied the correction method to simulated MEG data and real resting-state MEG recordings from eight subjects.
- Computed partial correlations between power envelopes of corrected ROIs to infer functional connectivity.
Main Results:
- The symmetric orthogonalisation method accurately reconstructed simulated functional networks.
- Analysis of real MEG resting-state data revealed dense bilateral connections within motor and visual networks.
- Identified direct, longer-range fronto-parietal connections in the human brain's functional network.
Conclusions:
- Symmetric orthogonalisation effectively corrects for spurious correlations in MEG source reconstruction.
- The corrected ROIs facilitate reliable application of network modeling techniques for connectome inference.
- This method enhances the understanding of functional brain connectivity patterns from MEG data.
