EEG vs. MEG: Same Brain, Different Physics

The Same Neurons, Two Completely Different Detectors

Here's something that should stop you in your tracks for a second.

Every time a neuron in your brain fires, it does two things simultaneously. It creates a tiny electrical field. And it creates a tiny magnetic field. Same neuron. Same moment. Two completely different physical phenomena rippling outward from the same source.

We built two entirely separate technologies to detect each one. Electroencephalography (EEG) picks up the electrical fields. Magnetoencephalography (MEG) picks up the magnetic fields. They're reading the same page of the same book, but one is reading it in English and the other in Japanese.

And yet these two technologies could not be more different in practice. One fits on your head like a pair of headphones and costs less than a decent laptop. The other weighs several tons, sits inside a room lined with layers of magnetically shielding metal, drinks liquid helium to stay at -269C, and costs more than most houses.

How did two technologies born from the same neural event end up living such wildly different lives? And which one should you actually care about?

That story starts with a little bit of physics. Don't worry. It's the fun kind.

How Your Brain Makes Electricity and Magnetism at the Same Time

To understand why EEG and MEG exist as separate technologies, you need to understand what happens when a neuron fires.

Your brain contains roughly 86 billion neurons. When a neuron receives a signal from its neighbors, ions (charged particles) flow across its cell membrane through tiny channels. Sodium ions rush in. Potassium ions rush out. This creates a flow of electrical current along the neuron.

Now, here's where physics gets beautiful. Whenever electrical current flows, it creates two things. First, it creates an electrical potential, a voltage difference that spreads through the surrounding tissue. Second, by the laws of electromagnetism (thank you, James Clerk Maxwell), that same current generates a magnetic field that wraps around the current like an invisible sleeve.

One event. Two signals. Both carrying information about what that neuron just did.

A single neuron's signals are absurdly weak. The electrical potential from one neuron is measured in microvolts, about a million times weaker than a AA battery. The magnetic field is measured in femtoteslas, roughly a billion times weaker than the Earth's magnetic field.

But neurons don't work alone. When tens of thousands of pyramidal neurons in the cortex fire in synchrony, their individual signals add up. The electrical potentials sum together into waves strong enough to detect through the skull, skin, and hair on your head. The magnetic fields sum too, producing a field just barely strong enough to detect with extraordinarily sensitive instruments.

This is the fork in the road. The electrical signal and the magnetic signal take very different paths from brain to sensor, and those paths determine everything about EEG and MEG.

EEG: Reading Your Brain's Electrical Chatter

EEG was the first to arrive. Hans Berger, a German psychiatrist, recorded the first human EEG in 1924 using silver wires inserted under his patient's scalp (ouch) and a sensitive galvanometer. He discovered alpha brainwaves, those smooth 8-13 Hz oscillations that appear when you close your eyes and relax, and spent years convincing a skeptical scientific community that he was actually recording brain activity and not muscle artifacts.

The basic principle hasn't changed in a century. Place electrodes on the scalp. Measure voltage differences between them. Those voltage patterns reflect the synchronized electrical activity of millions of cortical neurons firing below.

But here's the thing about electrical signals and skulls: the skull is a terrible conductor. When electrical fields pass through bone, cerebrospinal fluid, and skin, they get smeared. It's like looking at city lights through frosted glass. You can see that there's activity, and you can see the overall patterns, but the fine spatial details are blurred.

This is why EEG has excellent temporal resolution (it can track changes millisecond by millisecond) but limited spatial resolution (it can't pinpoint exactly where in the brain a signal originates with high precision). The electrical signals arrive fast, but they arrive blurred.

The beauty of EEG is its simplicity and accessibility. The electrodes are just conductive sensors pressed against your scalp. No surgery. No special room. No cryogenic cooling. Modern consumer EEG devices have shrunk the technology from a full clinical lab setup down to something you can wear while working at your desk or meditating on your couch.

MEG: Listening to Your Brain's Magnetic Whisper

MEG came along much later. In 1968, David Cohen at MIT used a copper induction coil inside a magnetically shielded room to detect the magnetic fields produced by alpha waves in a human brain. The signal was barely distinguishable from noise. Then, in 1972, he repeated the experiment using a brand-new kind of sensor called a SQUID (Superconducting Quantum Interference Device), and the signals jumped into sharp relief.

SQUIDs are the most sensitive magnetic field detectors ever built. They work by exploiting a quantum mechanical effect in superconducting loops cooled to near absolute zero. At -269C (just 4 degrees above absolute zero), certain materials become superconductors, carrying electrical current with zero resistance. A SQUID uses this property to detect magnetic fields so faint that they're a billion times weaker than a refrigerator magnet.

This is why MEG machines are so large and expensive. You're not just building a brain scanner. You're building a system that maintains sensors at near absolute zero temperature, continuously fed by liquid helium that costs tens of thousands of dollars per year to replenish. And you need to put the whole thing inside a magnetically shielded room, because without shielding, the Earth's magnetic field and every electronic device in the building would overwhelm the brain's tiny magnetic signals like a jet engine drowning out a whisper.

But here's why it's worth all that trouble: magnetic fields pass through the skull virtually undistorted.

Think about that for a second. The skull, which smears EEG's electrical signals into blurry spatial maps, is essentially transparent to MEG's magnetic signals. The magnetic field that reaches MEG's sensors looks almost exactly like the magnetic field that left the brain. No smearing. No distortion. The frosted glass is gone.

This gives MEG a significant advantage in source localization, the ability to pinpoint where in the brain a signal is coming from. MEG can typically localize neural sources to within 2-3 millimeters, compared to EEG's roughly 1-2 centimeters. Both technologies share millisecond-level temporal resolution, but MEG paints a sharper spatial picture.

The Head-to-Head Comparison

So we have two technologies born from the same neural event, diverging wildly in practice. Let's put them side by side.

The numbers tell a clear story. These technologies share their greatest strength (millisecond temporal resolution) but diverge on almost everything else. MEG wins on spatial precision. EEG wins on literally everything related to practical use.

But the comparison gets more interesting when you look at what each technology can and can't see.

The Blind Spots: What Each One Misses

Here's something most comparison articles won't tell you. EEG and MEG don't just differ in resolution. They actually have different blind spots. They see different subsets of the same brain activity.

Remember those pyramidal neurons arranged in columns in the cortex? The cortex isn't smooth. It's folded into ridges (gyri) and grooves (sulci), like a crumpled piece of paper. The orientation of neural columns relative to the scalp surface depends on where they sit on this crumpled surface.

Neurons on the tops of the ridges (gyri) are oriented radially, pointing straight out toward the scalp. Their electrical fields project directly outward and are easily detected by EEG. But their magnetic fields form loops parallel to the scalp surface, making them nearly invisible to MEG sensors positioned above the head.

Neurons on the walls of the grooves (sulci) are oriented tangentially, running parallel to the scalp surface. Their magnetic fields project outward and upward, making them easy targets for MEG. But their electrical contributions to the scalp surface are relatively weaker and more diffuse.

This is a genuinely fascinating result. EEG and MEG, reading the same brain at the same time, are partially blind to different populations of neurons based purely on geometry. EEG sees both radial and tangential sources (though tangential sources are somewhat attenuated). MEG primarily sees tangential sources and is nearly blind to radial ones.

This is why the best neuroscience studies use both. Together, they cover each other's blind spots. Separately, each one gives you a systematically incomplete picture. About two-thirds of cortical surface area lies in the sulci, meaning MEG captures the majority of the cortex's tangential activity quite well. But that remaining third on the gyri? EEG has it covered.

The Neurosity Crown gives you real-time access to your own brainwave data across 8 EEG channels at 256Hz, with on-device processing and open SDKs.

See the Crown

When Researchers Choose MEG

Given MEG's cost and complexity, why would anyone use it? Because for certain research questions, the spatial precision is worth every penny and every liter of liquid helium.

Pre-surgical mapping for epilepsy. This is MEG's clinical crown jewel. Before a neurosurgeon removes brain tissue to treat drug-resistant epilepsy, they need to know exactly where the seizure focus is and exactly where critical functions like language and motor control live. MEG's source localization accuracy can mean the difference between removing the right tissue and causing irreversible damage. It's one of the few clinical applications where MEG is considered a standard, reimbursable procedure.

Auditory and somatosensory research. MEG excels at studying how the brain processes sounds and touch. The primary auditory and somatosensory cortices sit along the walls of sulci, making them tangentially oriented and perfectly suited for MEG detection. Some of the most detailed maps of auditory processing in the human brain come from MEG studies.

Language processing studies. Understanding how the brain processes language in real time, word by word, millisecond by millisecond, requires both the temporal precision to track rapid processing and the spatial precision to distinguish between nearby language areas. MEG provides both. It's been instrumental in mapping the timing of syntactic and semantic processing across cortical regions.

Connectivity and oscillation research. When neuroscientists want to study how different brain regions communicate through synchronized oscillations, MEG's clean spatial signal makes it easier to determine which regions are actually talking to each other versus which ones just appear connected because of spatial smearing.

When EEG Is the Clear Winner

For most applications outside those specialized research niches, EEG isn't just a reasonable alternative to MEG. It's the better tool.

Sleep studies. You can't sleep in an MEG machine (any head movement corrupts the data). EEG is the backbone of sleep research and clinical sleep medicine. Polysomnography, the gold standard sleep test, centers on EEG recording. People sleep in EEG for entire nights without issue.

Long-duration monitoring. Clinical EEG can monitor patients continuously for days, watching for seizures in an epilepsy monitoring unit. MEG sessions rarely exceed an hour because subjects must remain motionless and the liquid helium supply is finite.

Brain-computer interfaces. This is where the accessibility gap becomes a canyon. BCIs need real-time brain data from a system the user can actually wear while doing things. A speller BCI that requires a magnetically shielded room and a team of technicians defeats the purpose. EEG's portability isn't just convenient here. It's essential.

Neurofeedback and cognitive training. Giving your brain real-time feedback on its own activity requires a wearable system that works in natural environments. EEG makes this possible anywhere: at home, in a clinic, at your desk. MEG makes it possible only in a very expensive room.

Population-scale research. When you need to study thousands of subjects (as in genetic studies of brain function or large-scale clinical trials), the $500-$1,500 per session cost of MEG versus the near-zero marginal cost of consumer EEG makes the choice obvious.

Everything involving movement. MEG requires subjects to keep their head extremely still relative to the sensor helmet. EEG works while you walk, exercise, meditate, commute, or do pretty much anything. Modern consumer EEG devices include accelerometers to account for movement artifacts, making them strong in real-world conditions.

The Next-Generation Plot Twist: OPM-MEG

There's a new chapter being written in this story, and it's worth knowing about.

Traditional MEG uses SQUID sensors that need liquid helium cooling. But over the past decade, a new type of magnetic sensor has emerged: the Optically Pumped Magnetometer, or OPM. These sensors detect magnetic fields using the quantum properties of alkali atoms (usually rubidium) held in a small glass cell. They're sensitive enough to detect brain signals. And crucially, they operate at room temperature.

This changes things. Without the need for a cryogenic dewar, OPM-MEG sensors can be mounted in a lightweight helmet that sits directly on the subject's head. The sensors move with the head, so subjects can make natural movements. The systems are smaller, lighter, and considerably cheaper than traditional MEG.

But don't get too excited just yet. OPM-MEG systems still require magnetic shielding (the sensors are so sensitive that the Earth's magnetic field would saturate them). The shielded rooms are smaller and cheaper than traditional MEG rooms, but they still cost hundreds of thousands of dollars. And while the per-sensor cost is dropping, a full-head OPM-MEG system still runs well into six figures.

OPM-MEG is a genuine breakthrough for MEG research. It's making MEG possible for populations that couldn't use it before, like children (who can't hold still for traditional MEG) and patients with movement disorders. But it's not about to show up at your local electronics store. The accessibility gap between EEG and MEG has narrowed, but it remains enormous.

Why Accessibility Wins in the End

Here's the thing that gets lost in technical comparisons of spatial resolution and source localization accuracy. The most important feature of any brain measurement technology is whether anyone can actually use it.

MEG can resolve neural sources to within a few millimeters. That's remarkable. But what good is millimeter precision if the technology is locked inside 300 facilities worldwide? There are roughly 8 billion brains on this planet. Fewer than a few thousand people per year get an MEG scan.

EEG, by contrast, is everywhere. It's in hospitals, clinics, research labs, and, increasingly, in people's homes. Consumer EEG devices have put real-time brainwave measurement into the hands of hundreds of thousands of people. And each of those people can generate data every single day, for months or years, building longitudinal datasets that would be physically impossible with MEG.

This is the pattern we see again and again in technology. The tool with the highest raw performance doesn't win. The tool that gets into the most hands wins. CT scanners didn't replace X-rays. Professional cinema cameras didn't replace smartphone cameras. The "good enough" technology that everyone can access generates more knowledge, more innovation, and more impact than the perfect technology that sits behind a locked door.

The Neurosity Crown is built on this exact philosophy. Eight EEG channels sampling at 256Hz, with on-device processing via the N3 chipset, in a form factor that weighs 228 grams and fits on your head like a pair of headphones. It doesn't have millimeter spatial resolution. What it has is something more powerful: it's there. On your desk. Ready when you are. No appointment, no technician, no shielded room.

And with open developer tools, the Crown turns brainwave data into something you can build with. JavaScript and Python SDKs let you create applications that respond to your brain state in real time. The MCP integration lets your brain data talk directly to AI tools like Claude. Try doing that with an MEG machine.

The spatial resolution difference between EEG and MEG matters in a neurosurgery planning session. It does not matter when you're trying to understand your own focus patterns, train your brain through neurofeedback, or build the next generation of brain-powered applications.

The Future Is Both (But Mostly EEG)

The honest answer to "EEG vs MEG: which should I use?" is that they're not really competitors. They're complementary tools designed for different contexts.

If you're a neurosurgeon planning an epilepsy resection, use MEG (and EEG, and fMRI, and everything else you can get). If you're a cognitive neuroscientist studying millisecond-level dynamics of language processing with high spatial precision, MEG is probably worth the investment. If you're investigating how brain regions coordinate through oscillatory coupling and you need clean source separation, MEG gives you an edge.

For virtually everything else, EEG is the answer. Sleep research. BCI development. Neurofeedback. Cognitive monitoring. Everyday brain awareness. Population studies. Remote research. And the entire emerging field of personal neurotechnology.

The most exciting brain science of the next decade won't come from bigger, more expensive machines in more shielded rooms. It will come from smaller, more accessible sensors on more heads, generating more data, in more real-world contexts. The brain doesn't just exist in a laboratory. It exists while you're thinking through a hard problem at 2am, while you're meditating before a big presentation, while you're writing code and trying to hold an entire system architecture in your working memory.

Those are the moments that matter. And those are the moments that only EEG can capture.

Your brain is producing electrical and magnetic fields right now, as you read this sentence. One of those signals requires a multi-ton machine in a shielded vault to detect. The other one? You could be reading it with a device on your head before you finish your next cup of coffee.

The physics is the same. The accessibility could not be more different. And accessibility, it turns out, is everything.