Glossary
Magnetoencephalography
What is magnetoencephalography?
Magnetoencephalography (MEG) records weak magnetic fields generated by neural currents. It offers millisecond timing and can estimate source locations under suitable recording and modeling conditions.
How it works
Conventional systems use superconducting quantum interference devices (SQUIDs) as sensors. They require cryogenic cooling, typically supplied by liquid helium, and recordings commonly take place in a magnetically shielded room to reduce interference from the environment. Newer optically pumped magnetometer systems can record without cryogenic cooling. Localization depends on sensor configuration, noise, source geometry and analysis; MEG does not guarantee better localization than EEG in every setting.
Applied example
A research team might present a stimulus, record the response, and estimate where the signal arose. In presurgical epilepsy assessment, a team can estimate the sources of abnormal discharges and compare them with the patient’s MRI and other recordings. These candidate locations can help guide placement of electrodes for further testing of possible seizure-onset sites. MEG can also help investigate language-related activity. Such estimates contribute to planning; they do not by themselves define safe surgical boundaries or map every function that must be preserved.
Why it matters
MEG helps relate the timing of neural signals to estimated sources. Different systems have different costs, movement constraints and technical requirements.
Evidence note: Schofield and colleagues (2024) demonstrate an OPM-MEG system that does not require cryogenic cooling.