Magnetoencephalography

Magnetoencephalography (MEG) is a noninvasive neuroimaging technique that records the magnetic fields produced by neuronal electrical activity in the brain. The measured signals primarily reflect synchronized postsynaptic currents in cortical pyramidal neurons arranged perpendicular to the cortical surface, with field strengths on the order of femtoteslas. MEG provides direct millisecond-precision measurements of brain activity.

MEG systems use ultra-sensitive magnetometers, historically superconducting quantum interference devices (SQUIDs), housed in magnetically shielded rooms to reduce ambient noise. More recently, room-temperature optically pumped magnetometers have been developed. Recordings are complemented by anatomical MRI to allow precise localization of functional activity within the brain.

Analysis of MEG data typically involves separating time-locked or oscillatory signals from background activity. Evoked magnetic fields reflect responses to specific stimuli, while induced activity and resting-state oscillations reveal ongoing brain dynamics. Source localization employs mathematical models of the head and inverse methods to estimate where in the brain the signals originate.

Applications of MEG include functional mapping for pre-surgical planning, particularly to localize epileptogenic zones, language and sensory processing areas, and the motor cortex. It is also used in cognitive and developmental research to study perception, attention, memory, and resting-state networks. MEG is often combined with MRI and EEG to enhance spatial localization and interpretability.

Advantages of MEG include excellent temporal resolution and noninvasiveness, with good sensitivity to cortical activity. Limitations include relatively limited ability to detect deep brain sources, the nonuniqueness of the inverse problem, high cost, and the need for shielded environments and specialized analysis. Safety is noninvasive and generally considered very high.