Magnetoencephalography can reveal deep brain network activities linked to memory processes

Area of Science:

• Neuroscience
• Biophysics
• Cognitive Science

Background:

• Noninvasive recording of deep neural structures like the hippocampus with high temporal resolution is a significant challenge in human neuroscience.
• Magnetoencephalography (MEG) potentially offers a solution, but deep structure signals may be too weak or masked by neocortical activity.

Purpose of the Study:

• To demonstrate that deep brain structures, specifically the hippocampus, can be noninvasively recorded using magnetoencephalography (MEG).
• To validate the use of blind source separation (BSS) techniques for isolating deep neural activity from MEG signals.
• To explore the potential of MEG for studying deep brain structures in cognitive processes and neurological disorders.

Main Methods:

• Applied blind source separation (BSS) to magnetoencephalography (MEG) signals recorded during a memory recognition task.
• Simultaneously recorded intracranial electroencephalography (EEG) signals from the hippocampus and rhinal cortex in epilepsy patients.
• Validated the isolated MEG components by correlating their time courses with simultaneously recorded stereo-EEG signals.

Main Results:

• A distinct mesial brain component was successfully identified in MEG signals from all participants using BSS.
• The time course of this mesial component showed selective correlation with hippocampal and rhinal cortex stereo-EEG signals.
• This confirms the mesial origin of the identified MEG component.

Conclusions:

• Blind source separation (BSS) enables noninvasive magnetoencephalography (MEG) to record activity from deep brain structures like the hippocampus.
• This technique successfully disentangles deep mesial activity from widespread neocortical networks.
• This finding opens new avenues for using MEG to investigate deep brain function in cognition and disease.