MEG

MEG
Invented by	David Cohen
Year	1968
Manufacturers	Yokogawa, Elekta, EKKO
Application	Neurology, Neurosurgery, Cognitive neuroscience

MEG

Magnetoencephalography is a noninvasive neurophysiological technique that records magnetic fields produced by neuronal electrical currents. It provides millisecond-scale temporal resolution and spatial localization of cortical activity used in clinical care and cognitive research. Developed from advances in cryogenic sensor technology and signal processing, it complements imaging modalities used in neurology and neuroscience.

Introduction

Magnetoencephalography measures extracranial magnetic fields arising from synchronous postsynaptic currents in populations of pyramidal neurons. Invented by David Cohen and advanced with contributions from groups at University of California, Berkeley, MIT, and University of Nottingham, the method became practical after the development of the superconducting quantum interference device by John Bardeen-inspired teams and industrial partners such as Elekta. Clinical adoption began in centers including Mayo Clinic, Johns Hopkins Hospital, and Great Ormond Street Hospital for presurgical mapping and epilepsy evaluation. Prominent research programs at Massachusetts General Hospital, Max Planck Society, and University College London use it alongside electrophysiology, structural imaging, and neuromodulation studies.

Principles and Technology

Sensors detect femtotesla-scale magnetic fields generated primarily by intracellular currents aligned perpendicular to the cortical surface in cortical pyramidal cells. Core hardware elements include cryogenic magnetometers such as superconducting quantum interference devices developed with influence from Harvard University and Bell Labs, magnetically shielded rooms produced by companies collaborating with NIST, and head localization systems compatible with structural imaging from Siemens Healthineers and GE Healthcare. Signal chains use differential amplification and digital sampling techniques refined at laboratories like California Institute of Technology and Stanford University. Source modeling and inverse solutions draw on mathematical methods advanced at University of Cambridge, ETH Zurich, and University of Oxford.

Clinical and Research Applications

Clinical applications emphasize localization of epileptogenic zones in drug-resistant focal epilepsy referred to specialized centers such as Cleveland Clinic and Charité – Universitätsmedizin Berlin. Preoperative functional mapping for language and sensorimotor cortex supports neurosurgical planning at institutions like University of Pennsylvania Health System and Mount Sinai Health System. Research applications span cognitive paradigms studied at Columbia University, auditory processing investigations linked to Carnegie Mellon University, studies of oscillatory dynamics in disorders examined by teams at McGill University, and developmental neuroscience at University of Toronto. MEG contributes to populations studied in translational work on Alzheimer's disease, Parkinson's disease, Schizophrenia, Autism spectrum disorder, and Stroke rehabilitation research.

Data Acquisition and Analysis

Acquisition protocols employ evoked and induced paradigms synchronized to stimuli used in experiments at facilities such as UCL Institute of Neurology and Riken laboratories. Core preprocessing includes artifact rejection methods developed in collaborations with MIT and Brown University to remove cardiac and ocular artifacts, and head-movement compensation techniques from Columbia University. Source reconstruction algorithms—beamformers, minimum norm estimates, and dipole fitting—were refined in groups at University of Helsinki, Karolinska Institutet, and Donders Institute. Time–frequency decomposition, connectivity metrics, and graph-theory analyses derive from computational neuroscience groups at Princeton University, University of Chicago, and Université de Montréal.

Comparison with Other Neuroimaging Modalities

Compared with modalities such as functional magnetic resonance imaging, MEG offers superior temporal resolution and direct sensitivity to neuronal currents but less sensitivity to deep structures like the thalamus compared with invasive recordings at centers such as NIH Clinical Center. Relative to electroencephalography, MEG is less distorted by skull conductivity variations and often provides improved spatial specificity for cortical sources, a distinction emphasized by researchers at Yale University and Albert Einstein College of Medicine. When combined with structural scans from University of California, San Francisco or metabolic imaging from Johns Hopkins University School of Medicine, multimodal approaches enhance localization and interpretation for clinical teams at institutions like UCLA Health.

Limitations and Challenges

Technical limitations include high acquisition cost driven by cryogenics and shielded facilities procured through vendors such as Siemens Healthineers and Elekta, limiting availability to tertiary centers like Toronto Western Hospital. Sensitivity decreases for deep or radially oriented sources described in studies from University of Wisconsin–Madison and McMaster University. Analysis faces ill-posed inverse problems addressed by mathematical work at Imperial College London and validation challenges requiring multimodal corroboration with intracranial electroencephalography in referral centers like University Hospitals Leuven. Standardization and normative databases remain active areas at consortia including Human Brain Project and multicenter initiatives led by European Union research programs.

History and Development

Early demonstrations of magnetic signals from the human brain by David Cohen in the 1960s were limited by sensor noise until the introduction of SQUID sensors and magnetically shielded rooms in the 1970s and 1980s at institutions such as MIT and Stanford University. Pioneering clinical and methodological expansions occurred through collaborative networks involving Mayo Clinic, Massachusetts General Hospital, and University College London during the 1990s and 2000s, paralleled by industrial development from Elekta and academic toolboxes from OHBA-affiliated groups. Recent trends include wearable optically pumped magnetometers developed with partnerships at University of Nottingham and multinational projects supported by Wellcome Trust and national research agencies.

Category:Neuroimaging instruments