What Is Transcranial Magnetic Stimulation?

A Physicist, a Coil of Wire, and an Involuntary Thumb Twitch

In 1985, Anthony Barker walked into a lab at the University of Sheffield carrying a device that looked like it had no business being near a human head. A large, flat coil of wire connected to a power supply capable of discharging thousands of amps in a fraction of a second.

He held the coil against a volunteer's scalp, right over the motor cortex. He fired it.

The volunteer's thumb twitched. Not because the volunteer wanted it to. Not because anyone touched their hand. The magnetic field passed straight through the skull, induced an electrical current in the motor neurons beneath, and those neurons fired whether they liked it or not.

The volunteer felt a clicking sensation. A tap on the scalp. Nothing painful. And yet something extraordinary had just happened: for the first time, a scientist had reached into a living human brain and made specific neurons fire, without surgery, without electrodes, without breaking the skin.

That device was the first transcranial magnetic stimulator. And the principle it demonstrated, that a magnetic field can pass through bone and force neurons to activate, would eventually transform psychiatry, neurology, and our understanding of how the brain is wired.

The Physics Is Simpler Than You Think

Here's the thing about TMS that surprises most people: the underlying physics is not complicated. A motivated high school student could understand it. Michael Faraday worked out the key principle in 1831, more than 150 years before anyone thought to point it at a brain.

Faraday's discovery was this: a changing magnetic field creates an electric current in any conductor nearby. Run electricity through a coil of wire and you get a magnetic field. Change that magnetic field rapidly and it will induce current in anything conductive within range. This is electromagnetic induction. It's how generators work, how wireless chargers work, and how your electric stovetop heats a pan.

Your brain happens to be conductive. Neurons communicate using electrical signals. The fluid surrounding them is salty and conducts electricity. So when you hold a coil against someone's head and pulse a rapidly changing magnetic field through it, that field passes through the skull (bone doesn't block magnetism the way it blocks electricity) and induces a small electrical current in the cortical tissue underneath.

That induced current is enough to depolarize neurons. To push them past their firing threshold. To make them send signals to other neurons, which fire in turn, cascading through whatever circuit they're connected to.

Think about it this way. Your brain runs on electricity. TMS is a way to inject electricity into a specific spot without ever touching the brain itself. The magnetic field is just the delivery vehicle, the thing that gets the electricity through the skull.

What One Pulse Can Tell You

Before TMS became a treatment, it was a diagnostic tool. And in many ways, the diagnostic applications are even more elegant than the therapeutic ones.

When Barker first zapped that volunteer's motor cortex and watched the thumb twitch, he wasn't trying to treat anything. He was trying to measure something. Specifically, he wanted to measure how quickly a signal travels from the brain down to the muscles.

Here's how it works. You place the TMS coil over the motor cortex, right above the strip of brain tissue that controls movement. You fire a single pulse. Somewhere in the body, a muscle twitches. You record the twitch with an electromyography (EMG) sensor on the skin over the muscle. The time between the pulse and the twitch tells you how fast the signal traveled down the nerve pathway.

This measurement, called the motor evoked potential (MEP), turned out to be incredibly useful. In healthy people, the signal travels at a predictable speed. In people with multiple sclerosis, the myelin sheath insulating the nerve fibers is damaged, and the signal slows down. In people with spinal cord injuries, the signal might not arrive at all.

Single-pulse TMS became a standard clinical tool for assessing the integrity of motor pathways. Neurologists use it to help diagnose conditions affecting the corticospinal tract, to monitor surgical patients during operations near the spinal cord, and to map which parts of the motor cortex control which muscles.

That motor mapping capability is worth pausing on. By moving the TMS coil across the scalp and firing single pulses, a clinician can build a map of which brain regions control which body parts. Stimulate here and the thumb twitches. Move the coil two centimeters and now it's the index finger. Move again and it's the wrist.

This is essentially the same motor homunculus that Wilder Penfield mapped in the 1930s by directly stimulating the exposed brains of awake surgical patients. TMS lets you do it without opening the skull.

From Single Pulses to Treatment: The Birth of rTMS

Single-pulse TMS tells you what the brain is doing right now. But the real clinical story starts when you fire pulses repeatedly. This is called repetitive TMS, or rTMS. And it does something that single pulses can't: it produces changes in brain activity that outlast the stimulation itself.

The mechanism behind this is related to a fundamental property of neurons called plasticity. Neurons that fire together repeatedly strengthen their connections. Neurons that fire out of sync weaken theirs. This is the basis of all learning and memory, summarized in the famous phrase "neurons that fire together, wire together."

When you deliver rTMS to a brain region, you're essentially forcing a population of neurons to fire in a specific temporal pattern over and over. Do this enough times and the excitability of that neural population changes. Not just during the stimulation, but for minutes, hours, or (with repeated sessions over days and weeks) potentially months afterward.

The direction of the change depends on the frequency:

This frequency-dependent control is what makes rTMS a precision tool. If brain imaging shows that a particular region is underactive in a patient, you can use high-frequency stimulation to boost it. If a region is overactive, you can use low-frequency stimulation to calm it down.

The Depression Breakthrough That Changed Psychiatry

The biggest clinical story in TMS, the one that put it on the map and into thousands of psychiatric offices, is its use for major depressive disorder.

The logic goes like this. Decades of neuroimaging research have shown that people with depression tend to have reduced activity in the left dorsolateral prefrontal cortex (left DLPFC). This region is involved in positive emotional processing, executive function, and the ability to regulate the darker signals coming from deeper limbic structures like the amygdala.

In depression, the left DLPFC is underperforming. It's not doing its job of putting the brakes on negative emotion. The hypothesis was elegant: use high-frequency rTMS to wake up the left DLPFC, restore its activity, and the depressive symptoms should improve.

Researchers tested this throughout the 1990s and 2000s. The results were consistent enough that in 2008, the FDA cleared the first rTMS system (Neuronetics' NeuroStar) for adults with major depressive disorder who had failed to respond to at least one antidepressant medication.

In the pivotal trial, patients received daily TMS sessions for 4 to 6 weeks. About 14% achieved full remission. That number might sound modest, but remember the context: these were treatment-resistant patients. They had already tried medication and it hadn't worked. For them, 14% remission was a meaningful signal.

Since then, the numbers have gotten better. A lot better.

What TMS Can (and Can't) Reach

There's a physical limitation to TMS that defines its boundaries, and understanding it helps explain both its power and its constraints.

The magnetic field generated by a standard figure-eight TMS coil drops off rapidly with distance. It can effectively stimulate cortical tissue to a depth of about 1.5 to 3 centimeters from the scalp surface. That's enough to reach the outer layer of the brain, the cortex, but not enough to directly reach the deeper structures underneath.

This matters because many of the brain circuits involved in psychiatric conditions extend well below the cortical surface. The amygdala, hippocampus, nucleus accumbens, subgenual cingulate cortex: these structures sit deep inside the brain, well beyond the reach of a standard TMS coil.

So how does TMS for depression work if it can't directly reach the limbic regions involved in mood? The answer is connectivity. The left DLPFC has strong anatomical connections to those deeper structures. Stimulate the surface node of the circuit, and the effects propagate downward through the network. It's like knocking on the front door to reach someone in the basement. You can't get to them directly, but the sound travels.

Deep TMS (dTMS) attempts to address this limitation with a different coil design. Instead of a figure-eight, deep TMS uses an H-coil, a more complex geometry that generates magnetic fields reaching 3 to 4 centimeters deep. It sacrifices some spatial precision for greater depth of penetration. The FDA cleared a deep TMS device (BrainsWay) for depression in 2013 and for OCD in 2018.

Beyond Depression: The Expanding Clinical Map

Depression was the proving ground, but researchers have been exploring TMS across a wide range of conditions. Here's where the evidence stands as of 2026:

The pattern is interesting. TMS works best when there's a clear, identifiable brain circuit that's malfunctioning in a predictable direction (too active or not active enough) and when that circuit is close enough to the cortical surface to be reached by the coil, either directly or through well-established connections.

Brainwave data, captured at 256Hz across 8 channels, processed on-device. The Crown's open SDKs let developers build brain-responsive applications.

Explore the Crown

The Research Tool: TMS as a Window Into Causation

There's a side of TMS that gets less public attention than the clinical applications, but that neuroscientists arguably value even more: its use as a research tool.

Here's the problem TMS solves for researchers. Most brain imaging techniques, including EEG, fMRI, and fNIRS, are correlational. They can tell you that a brain region is active during a particular task, but they can't tell you whether that region is actually necessary for the task. Maybe it's just along for the ride. Maybe it's involved in something else that happens to co-occur with the task.

TMS can answer the causation question. If you temporarily disrupt a brain region with TMS and the person's performance on a specific task gets worse, you can conclude that the region is causally involved in that task. Not just correlated. Actually necessary.

This technique, sometimes called a "virtual lesion," has been enormously productive in cognitive neuroscience. Researchers have used it to establish causal roles for specific brain regions in language processing, visual perception, decision-making, working memory, and dozens of other cognitive functions.

And here's the part that connects TMS to EEG in a way that most people don't realize: some of the most powerful TMS research combines TMS with EEG recording.

This is called TMS-EEG. The idea is simple but the results are profound. You fire a TMS pulse at a specific brain region, and then you immediately record the resulting EEG signal to see how the rest of the brain responds. The TMS pulse is like dropping a stone in a pond. The EEG measures the ripples.

TMS-EEG reveals the functional connectivity of the brain. How does stimulating the prefrontal cortex affect activity in the parietal lobe? How does that connectivity pattern differ between healthy controls and patients with schizophrenia? How does it change after a course of antidepressant treatment?

These are questions you can't answer with either technique alone. TMS provides the causal perturbation. EEG provides the millisecond-resolution measurement of the brain's response. Together, they create a tool that is more than the sum of its parts.

The Practical Reality: What Getting TMS Actually Looks Like

If you've never seen a TMS session, the clinical reality might surprise you. It's more mundane than the science suggests.

You sit in a reclining chair, like a dentist's chair. A technician positions the TMS coil against your head, usually holding it in place with an articulated arm. For depression treatment, the coil goes over the left DLPFC, typically located by measuring from standard skull landmarks or, increasingly, by using neuronavigation software that maps the coil position onto a brain image.

The machine fires. It sounds like a rapid clicking or tapping. Each pulse lasts only a fraction of a millisecond, but during high-frequency stimulation, the pulses come fast: 10 per second for a standard protocol, or in rapid bursts for theta burst stimulation. You feel a tapping sensation on your scalp, sometimes described as a woodpecker. Some people find it mildly uncomfortable. Others barely notice it.

A standard session takes 20 to 40 minutes, depending on the protocol. Theta burst protocols can be as short as 3 minutes. You stay awake and alert the whole time. No anesthesia, no sedation, no recovery period. You walk out and drive yourself home.

The treatment course for depression involves coming back every weekday for 4 to 6 weeks. That's 20 to 30 sessions. The newer accelerated protocols (like the Stanford SAINT protocol) pack more sessions into fewer days, but they're not yet widely available outside research settings.

Cost is a real factor. A full treatment course runs $6,000 to $15,000. Insurance coverage is improving, especially for treatment-resistant depression, but many patients still face significant out-of-pocket expenses. This is not a casual investment.

And the effects aren't always permanent. Studies show that roughly half of initial responders experience some symptom recurrence within 6 to 12 months. Many clinicians recommend maintenance sessions, monthly or quarterly, to sustain the benefit. This means ongoing cost, ongoing clinic visits, and ongoing time commitment.

TMS and EEG: The Natural Partnership

Here's where the story takes a turn that might not be obvious.

TMS changes the brain. But how do you know what's actually changing? How do you track whether the treatment is working before the patient reports feeling different, which might take weeks? How do you personalize the treatment based on what a specific patient's brain actually needs?

This is where EEG enters the picture. Not as a competitor to TMS, but as its natural companion.

Before treatment, EEG can identify biomarkers that predict who will respond to TMS. Research has shown that specific patterns of frontal alpha asymmetry and prefrontal theta activity measured by EEG can help predict which patients are likely to benefit from rTMS for depression. A 2022 study in Brain Stimulation found that baseline EEG features predicted TMS treatment response with roughly 70% accuracy.

During treatment, TMS-EEG studies track how the brain's electrical response to stimulation changes over the course of treatment. Is the left DLPFC becoming more excitable? Is the connectivity between the DLPFC and deeper limbic regions strengthening? These changes show up in EEG data before they show up in clinical symptoms.

After treatment, ongoing EEG monitoring can detect early signs of relapse, changes in brainwave patterns that precede the return of depressive symptoms. This could allow clinicians to intervene with maintenance sessions before the patient experiences a full symptomatic relapse.

This is the major change happening in precision psychiatry: treating the brain not based on a one-size-fits-all protocol, but based on measurable neural data from each individual patient. And EEG is the most practical tool for collecting that data, because it's portable, affordable, and captures brain dynamics in real time.

Where Consumer EEG Fits In

TMS lives in the clinic. It requires specialized equipment, trained operators, and medical oversight. That isn't changing anytime soon, nor should it. Magnetic pulses powerful enough to make neurons fire are not something you want to DIY.

But the EEG side of the equation is a different story entirely.

Consumer EEG has reached a point where a device you wear like headphones can capture meaningful brain data with millisecond precision. The Neurosity Crown, for example, records from 8 EEG channels at 256Hz, covering frontal, central, and parietal-occipital regions. It measures the same electrical signals that clinical EEG systems measure, the same signals that TMS-EEG researchers use to track how stimulation changes the brain.

You can't do TMS at home. But you can do something that clinicians and researchers are increasingly recognizing as valuable: you can monitor your own brain's electrical patterns over time.

Track your frontal alpha asymmetry across days and weeks. Watch how your focus and calm scores change with sleep, exercise, caffeine, or meditation. Build a personal baseline of your own neural activity that provides context for any clinical intervention, whether that's TMS, neurofeedback, medication, or therapy.

The Crown's open JavaScript and Python SDKs mean that researchers and developers can build custom applications that bridge the gap between clinical brain stimulation and at-home brain monitoring. Imagine a future where your TMS clinician can review your at-home EEG data between sessions, adjusting protocol parameters based on real neural data rather than waiting for your next appointment to ask "How are you feeling?"

That future is not science fiction. The technology to make it happen exists today.

The Honest Limitations (Because They Matter)

TMS is not magic. It's a powerful, evidence-based tool with real limitations that deserve honest discussion.

We don't fully understand why it works. We know that rTMS changes neural excitability. We know it modulates connectivity between brain regions. But the precise mechanism connecting "magnetic pulse hits cortex" to "depressive symptoms improve" is still being worked out. The field has strong hypotheses and solid clinical data, but the mechanistic details are incomplete.

Response rates are imperfect. Even with the best protocols, not everyone responds to TMS. For standard rTMS in treatment-resistant depression, about 50-60% of patients show meaningful improvement and 30% achieve full remission. Those are meaningful numbers, but they also mean roughly half of patients don't get significant benefit.

Individual targeting is still evolving. The standard clinical approach positions the coil based on scalp measurements or basic structural MRI landmarks. The Stanford SAINT protocol showed that individualized targeting based on functional connectivity dramatically improves outcomes, but this approach requires functional brain imaging that most clinics don't yet offer.

Long-term data is still accumulating. TMS has been FDA-cleared since 2008, which gives us almost two decades of clinical experience. But compared to medications that have been studied for 50 or 60 years, the long-term picture is still filling in. What we know so far is reassuring: the safety profile remains strong, and the benefits are real, if sometimes temporary.

It can't reach everything. The cortical surface limitation means TMS is best suited for conditions involving superficial brain circuits. Conditions primarily driven by deep subcortical dysfunction may not respond as well, even with deep TMS coils that extend the reach somewhat.

What's Coming Next

The next generation of TMS is being shaped by three converging trends, and all of them point toward more personalized, more precise, and more data-driven treatment.

Individualized targeting using brain imaging and connectivity data. The SAINT protocol proved that aiming the coil at the exact right spot for each individual brain produces dramatically better outcomes. As functional brain imaging becomes more accessible and the algorithms for identifying optimal targets improve, this approach will become standard practice.

Closed-loop TMS guided by real-time brain data. The most exciting research frontier is closed-loop stimulation, where TMS pulses are triggered by real-time EEG or other brain measurements. Instead of delivering a fixed number of pulses to a fixed location on a fixed schedule, the system monitors the brain's state and delivers stimulation only when the brain is in the right state to benefit from it. Early studies suggest this approach could make TMS more effective while requiring fewer total pulses.

Integration with consumer brain monitoring. As consumer EEG devices provide increasingly reliable brain data, the gap between clinical TMS sessions and daily life is closing. At-home EEG monitoring between TMS sessions could provide the continuous data that clinicians need to optimize and personalize treatment in ways that periodic clinic visits never could.

The Bigger Picture

Zoom out far enough and TMS tells you something profound about the brain. It tells you that this organ, this three-pound universe of electrical and chemical activity, responds to physical intervention. That you can reach in with a magnetic field and change how it operates. That the patterns of neural activity underlying depression, OCD, chronic pain, and other conditions are not fixed. They can be shifted.

But it also tells you that the brain is not a passive recipient of intervention. It's an active, learning system. The reason TMS works is that the brain takes the externally imposed change and runs with it, reorganizing its own circuits in response.

This is the same principle that underlies neurofeedback, meditation, cognitive behavioral therapy, and every other approach to changing how the brain works. The tools are different. TMS uses magnetic pulses. Neurofeedback uses real-time EEG data. Meditation uses focused attention. But they all depend on the brain's fundamental capacity to reorganize itself in response to new information.

Understanding that capacity is the real prize. Not any single tool, but the growing ability to monitor, measure, and work with the brain's own plasticity. A world where you can watch your own brain's electrical patterns on a screen in real time, where clinicians can guide treatment based on neural data rather than symptom checklists, where the gap between "what your brain is doing" and "what you know about it" keeps shrinking.

That world is being built right now, one magnetic pulse and one brainwave measurement at a time.