Beamformer reconstruction of correlated sources using a modified source model.

Area of Science:

• Neuroimaging
• Biophysics
• Signal Processing

Background:

• Magnetoencephalography (MEG) analysis often uses beamformers to localize neural activity.
• Standard beamformers struggle to accurately reconstruct signals from multiple, temporally correlated sources.
• Accurate source localization is crucial for understanding complex brain dynamics.

Purpose of the Study:

• To introduce and validate a novel 'dual source beamformer' technique for MEG data.
• To demonstrate the dual source beamformer's ability to overcome the limitations of single source methods.
• To assess the technique's efficacy in imaging correlated neural activity, such as auditory responses.

Main Methods:

• Development of a lead field formulation for the dual source beamformer.
• Application of the dual source beamformer to simulated MEG data.
• Validation using simulations and real MEG data of auditory steady-state responses.

Main Results:

• The dual source beamformer accurately reconstructs power from two spatially distinct, temporally correlated sources.
• Standard single source beamformers significantly suppress the power of such correlated sources.
• Accurate imaging of correlated auditory steady-state responses between left and right auditory cortices was achieved.

Conclusions:

• The dual source beamformer offers a significant advancement for analyzing correlated neural activity in MEG.
• This technique is particularly effective when prior information about one source's location is available.
• It provides a valuable tool for neuroscientists studying synchronized brain function and inter-hemispheric communication.