Magnetoencephalography (MEG) is a brain imaging technique that detects the tiny magnetic fields produced by electrical activity in your neurons. It captures brain signals in real time, with millisecond precision, making it one of the fastest ways to watch the brain at work without surgery or radiation. MEG is used primarily in epilepsy care and presurgical brain mapping, and it’s gaining ground in research on conditions like autism, schizophrenia, and Parkinson’s disease.
How MEG Detects Brain Activity
Every time neurons in your brain fire, they produce a small electrical current. That current generates a magnetic field, but an extraordinarily weak one: roughly 100 femtotesla at the scalp. For perspective, Earth’s magnetic field is about a billion times stronger. Picking up a signal that faint requires specialized sensors and a carefully controlled environment.
Traditional MEG systems use sensors called SQUIDs (superconducting quantum interference devices). A SQUID is a superconducting loop of material interrupted by short sections of resistive material. When an external magnetic field hits the loop, the electrical current across those resistive sections shifts in a predictable way, turning the device into an extremely sensitive magnetic field detector. The catch is that superconductivity only works at temperatures near absolute zero, around negative 269°C. To stay that cold, the sensors are bathed in liquid helium, and the system typically burns through 10 to 12 liters of it per day.
Why MEG Needs a Shielded Room
Because the brain’s magnetic fields are so small, even background sources like power lines, passing cars, or nearby electronics would drown them out. MEG scans take place inside a magnetically shielded room (MSR), a structure built with multiple layers of specialized materials. The walls typically include several layers of MuMetal, a nickel-iron alloy with extremely high magnetic permeability that diverts low-frequency magnetic interference around the room instead of through it. A layer of copper or aluminum handles higher-frequency interference. Current commercially available rooms designed for the newest MEG sensors use four layers of MuMetal and one layer of copper.
Even with all that shielding, some residual magnetic field remains inside the room, typically around 10 to 30 nanotesla. Active shielding systems can further reduce this by generating corrective fields in real time.
Speed vs. Precision: How MEG Compares
MEG’s greatest strength is temporal resolution. It tracks brain activity on the order of milliseconds, meaning it can follow the rapid sequence of neural events as they happen. EEG shares this speed, but MEG typically localizes the source of that activity more accurately. In one study comparing the two, MEG localized brain signals to within about 10.8 millimeters of a reference point, while EEG averaged 16.6 millimeters of error. Combining the two brought that down to roughly 8.3 millimeters.
Functional MRI (fMRI), by contrast, offers millimeter-level spatial resolution, far sharper than MEG alone. But fMRI measures blood flow changes that lag seconds behind actual neural firing, so it can’t capture the timing of brain events. MEG and fMRI are often treated as complementary: fMRI shows precisely where something happened, MEG shows precisely when.
Clinical Uses in Epilepsy and Brain Surgery
The most established clinical application of MEG is in epilepsy. For patients whose seizures don’t respond to medication and who are being evaluated for surgery, MEG can help pinpoint the brain region where seizures originate. Its diagnostic accuracy for localizing seizure sources generally exceeds that of scalp EEG.
MEG is also used for presurgical functional brain mapping, identifying critical areas a neurosurgeon needs to avoid. Before an operation near the brain’s motor, language, or sensory regions, MEG can map exactly where those functions live in a specific patient’s brain. This matters because brain anatomy varies from person to person, and conditions like tumors or prior injury can shift functional areas away from their expected locations.
For language mapping specifically, MEG can determine which hemisphere of the brain is dominant for speech. This used to require the Wada test, an invasive procedure that involves temporarily anesthetizing one hemisphere at a time. Studies have shown that MEG language mapping can replace the language portion of that procedure entirely. MEG localizes receptive language processing to the posterior temporal lobe and expressive language to frontal and lower temporal areas, giving surgeons a functional roadmap before they operate.
Beyond language, MEG maps somatosensory cortex (the strip of brain that processes touch), primary motor cortex (movement), auditory cortex, and visual cortex. Each mapping uses a different task or stimulus during the scan: finger tapping for motor areas, tones for auditory cortex, visual patterns for the visual region.
What a MEG Scan Feels Like
MEG is completely passive. The machine does not emit any radiation, magnetic pulses, or energy of any kind. It simply listens for the magnetic fields your brain already produces. This makes it inherently safe, with no known biological risks from the scan itself.
The actual recording takes one to two hours, but you should expect to be at the facility for four to five hours total, accounting for preparation and any additional imaging. Before the scan, you’ll need to remove all jewelry, piercings, wigs, hair extensions, watches, and any other metal or electronic items. Even things you might not think about, like permanent makeup, tattooed eyeliner, or recent hair dye, can contain enough metallic particles to distort the signal.
During the scan, you sit with your head positioned inside a helmet-shaped sensor array. You’ll need to keep relatively still, since the rigid helmet in traditional systems doesn’t move with you. Depending on what’s being mapped, you may be asked to perform tasks: listening to words, tapping your fingers, or watching a screen.
Contraindications That Affect Signal Quality
Because MEG is passive, the contraindications aren’t about safety in the way MRI contraindications are. Metal in or on your body won’t heat up or move. Instead, ferromagnetic materials create magnetic artifacts that can overwhelm the brain signals the machine is trying to detect. The closer the metal is to the sensors (which surround your head), the bigger the problem.
Cochlear implants and certain brain fluid shunts are firm exclusions because they sit right where the sensors are measuring. Cardiac pacemakers, neurostimulators, insulin pumps, and other implanted electronic devices also disqualify patients. Joint prostheses, surgical clips, dental retainers, orthodontic braces, and even non-removable piercings can interfere, though some of these depend on the material and location. Tattoos below the elbow, for example, rarely cause issues. Some facilities offer a brief screening test, about 10 minutes, to check whether borderline cases like certain dental work will actually create artifacts before committing to a full scan session.
One less obvious restriction: if you’ve had an MRI within the past two weeks, residual magnetization in any metal in your body may not have fully dissipated, which can affect MEG recordings.
The Shift Toward Wearable MEG Systems
The biggest limitation of traditional MEG is the hardware. The cryogenic cooling system is expensive to operate, the rigid helmet fits only one head size, and patients must stay still throughout the scan. These constraints make it especially difficult to scan young children, who can’t hold still for long periods and whose heads are far smaller than the adult-sized helmet.
A newer generation of sensors called optically pumped magnetometers (OPMs) is changing this. OPMs achieve sensitivity comparable to SQUIDs without cryogenic cooling. They’re small and lightweight enough to be mounted directly on a flexible, wearable helmet that conforms to the individual’s head. Because the sensors sit closer to the scalp, with no vacuum gap needed for thermal insulation, they capture stronger signals and offer better spatial resolution.
The most transformative difference is that OPM-MEG allows movement during scanning. Participants can move their heads freely, which opens up entirely new types of experiments and makes scanning babies, toddlers, and patients with movement disorders far more practical. OPM-MEG systems do still require a magnetically shielded room, and in fact they impose stricter shielding requirements than traditional systems because the sensors operate within a very narrow magnetic field range. But the elimination of liquid helium alone dramatically reduces operating costs and complexity.