TMS vs. Neurofeedback

Two Ways to Change a Brain (and Why the Difference Matters More Than You Think)

Somewhere in a psychiatrist's office right now, a person is sitting in a padded chair while a technician holds a figure-eight-shaped coil against their scalp. The machine clicks rapidly, like a woodpecker on a deadline. With each click, a focused magnetic pulse passes through the skull and forces a cluster of neurons in the left prefrontal cortex to fire. The person feels a tapping sensation. Maybe a mild headache. After 20 minutes, the session is over. They'll come back tomorrow for another round, and the day after that, and the day after that. Thirty-six sessions total. That's the standard protocol.

On the other side of the country, a different person is sitting at their desk wearing what looks like a sleek pair of headphones. A screen in front of them shows a real-time visualization of their own brainwave activity. When their brain produces the target pattern (elevated sensorimotor rhythm, reduced theta), a tone plays and a score ticks upward. When their brain drifts, the feedback fades. Nobody is forcing their neurons to do anything. Nobody is even touching their head. They're watching their own brain's electrical output and learning, session by session, to shift it.

Both of these people are trying to change their brain activity. Both have science backing what they're doing. But the approaches are so fundamentally different that comparing them is a bit like comparing a personal trainer to a surgeon. One teaches your body to do something new. The other physically intervenes to make it happen.

This is the core tension between neurofeedback and transcranial magnetic stimulation (TMS). And understanding it will change how you think about what it actually means to "treat" the brain.

The Magnetic Hammer: How TMS Works

To understand TMS, you need to remember one fact from high school physics: a changing magnetic field creates an electric current.

Michael Faraday figured this out in 1831. He wrapped a coil of wire, ran current through it, and showed that the resulting magnetic field could induce current in a nearby conductor. He was thinking about generators and motors. He was absolutely not thinking about brains.

But in 1985, Anthony Barker at the University of Sheffield realized something: if you hold a coil of wire near someone's head and pulse a strong, rapidly changing magnetic field through it, the magnetic field passes through the skull (bone is not a barrier to magnetism, unlike electricity) and induces a small electrical current in the cortex beneath the coil. That induced current is strong enough to make neurons fire.

This was a big deal. For the first time, you could stimulate specific brain regions non-invasively, without surgery, without electrodes, without opening the skull. You just held a coil against someone's head and their motor cortex would fire, causing their thumb to twitch involuntarily. It was eerie and fascinating and immediately useful for neurological diagnosis.

But the real clinical breakthrough came when researchers realized that repeated stimulation, pulse after pulse delivered to the same brain region over many sessions, could produce lasting changes in neural excitability. This is called repetitive TMS, or rTMS. And it turns out that the direction of the lasting change depends on the stimulation frequency.

High-frequency rTMS (typically 10 Hz or above) tends to increase neural excitability in the targeted region. It makes those neurons more likely to fire on their own after the stimulation ends.

Low-frequency rTMS (1 Hz or below) tends to decrease neural excitability. It calms things down.

This frequency-dependent effect is the entire basis of TMS as a clinical tool. If you can identify a brain region that's underactive in a particular condition, you stimulate it at high frequency to wake it up. If a region is overactive, you use low frequency to quiet it.

The Limits of the Magnetic Hammer

So TMS is powerful. It's FDA-cleared. It has strong evidence for depression and growing evidence for OCD, smoking cessation, and chronic pain. But it has some very real constraints that define who can use it and how.

It requires a clinic. TMS devices are large, expensive medical instruments. The coils need precise positioning, often guided by neuronavigation systems that map the coil's position relative to the patient's brain anatomy. A trained technician or clinician operates the device and monitors for adverse reactions. You cannot do TMS at home. Full stop.

It's expensive. A standard treatment course for depression runs 30-36 sessions over 6 weeks. The total cost ranges from $6,000 to $15,000, depending on the clinic and geographic location. Insurance coverage has improved since FDA clearance, but it's inconsistent. Many patients pay a significant portion out of pocket.

The effects aren't permanent. This is the part that gets less airtime. While TMS produces real, meaningful changes in neural excitability, those changes don't always last. A 2015 meta-analysis in the Journal of Clinical Psychiatry found that 50-60% of initial TMS responders experienced symptom recurrence within 6-12 months. Many patients need "maintenance" sessions (monthly or quarterly) to sustain the benefit.

Side effects are generally mild but real. Scalp pain at the stimulation site is reported by 20-40% of patients. Headaches occur in about 25% of sessions. The most serious risk is seizure, which is rare (roughly 1 in 10,000 sessions) but non-trivial, especially given that sessions happen daily for weeks.

It only reaches the cortical surface. The magnetic field drops off sharply with distance. Standard TMS coils can effectively stimulate cortex to a depth of about 1.5 to 3 centimeters. This means deep brain structures, including the subcortical circuits heavily implicated in depression, anxiety, and addiction, can't be directly targeted. The clinical effects on those circuits happen indirectly, through the connections between the stimulated cortex and deeper regions.

Here's the thing about TMS that rarely gets stated plainly: it's something done to your brain by someone else. The patient sits passively while a machine forces their neurons to fire in a particular pattern. The brain doesn't learn anything during TMS. It doesn't develop a new skill. It receives a stimulus, and the stimulus produces a temporary shift in excitability that, with enough repetition, can become semi-durable.

This isn't a criticism. For severe, treatment-resistant depression, TMS can be genuinely life-saving. When someone's brain is stuck in a dysfunctional pattern and they've exhausted other options, external intervention makes perfect sense. You don't teach someone to swim when they're drowning. You throw them a rope.

But what about everyone else? What about people who want to improve focus, reduce anxiety, or train their brain to regulate itself better, not because they're in crisis, but because they want to perform and feel better than their baseline?

That's a different problem. And it calls for a different approach.

The Mirror: How Neurofeedback Works

Neurofeedback starts from a completely different premise than TMS. Instead of forcing neurons to fire, it asks: what if you could show the brain what it's doing and let it figure out how to change?

The basic setup is simple. EEG sensors on your scalp pick up your brain's electrical activity in real time. A computer processes the signal and extracts the relevant features, typically the power in specific frequency bands like theta (4-8 Hz), alpha (8-13 Hz), SMR (12-15 Hz), and beta (13-30 Hz). The computer then provides feedback: a visual display, a sound, a game, a score. When your brain produces the desired pattern, you get a reward signal. When it doesn't, the reward stops.

That's it. No magnets. No external force. Just information flowing in a loop: brain produces signal, computer reads signal, computer shows signal to the person, person's brain adjusts based on what it sees, and the loop repeats hundreds of times per session.

This is operant conditioning applied to neural oscillations. B.F. Skinner showed in the 1930s that any behavior that's followed by a reward becomes more likely to recur. What Barry Sterman discovered in the 1960s (initially in cats, then in humans) is that the same principle applies to brainwave patterns. Your cortex can learn to produce specific electrical rhythms, even though you have no conscious awareness of what those rhythms are. Give the brain feedback about its own activity, and it adjusts. Not because you're willing it to change. Because operant conditioning works on any trainable system, and your cortex is the most trainable system in the known universe.

The Evidence for Neurofeedback

The research base for neurofeedback is substantial, though it's important to be honest about where it's strong and where it's still developing.

ADHD brain patterns: The strongest evidence. A 2009 meta-analysis by Arns et al. found large effect sizes for neurofeedback on inattention (0.81) and impulsivity (0.69), comparable to stimulant medication. A 2014 follow-up showed these effects persisted 6 months after training ended. The American Academy of Pediatrics has rated neurofeedback as a "Level 1, Best Support" intervention for ADHD, the same rating they give to medication.

Anxiety: Growing support. Alpha/theta neurofeedback training has shown promise for generalized anxiety, with several controlled studies showing reductions in both subjective anxiety measures and physiological arousal. A 2019 randomized controlled trial in NeuroRegulation found that frontal alpha asymmetry training significantly reduced anxiety symptoms compared to a sham control.

Depression: Promising but earlier-stage. Several studies have found that training to increase left frontal alpha asymmetry (the same asymmetry pattern associated with reduced depression) produces improvements in mood. But the controlled trial evidence is not yet as strong as what exists for TMS. This is an area where more research is needed.

Peak performance and focus in healthy adults. SMR and beta/theta ratio training have shown improvements in sustained attention, working memory, and executive function in non-clinical populations. A 2015 NeuroImage study found that SMR training improved both behavioral performance and resting-state brain patterns in healthy adults.

Epilepsy: The original application. Sterman's early work showed that SMR training raised seizure thresholds in both animals and humans. Multiple controlled studies have since confirmed this effect.

The honest summary: neurofeedback has strong evidence for ADHD and epilepsy, good evidence for anxiety and focus enhancement, and promising but developing evidence for depression and other conditions. The field has historically been plagued by small sample sizes and variable methodology, but the quality of research has improved markedly in the last decade.

The Head-to-Head: Every Dimension That Matters

Now let's put these two approaches side by side.

The Crown captures brainwave data at 256Hz across 8 channels. All processing happens on-device. Build with JavaScript or Python SDKs.

Explore the Crown

The "I Had No Idea" Part: Why These Two Methods Might Actually Need Each Other

Here's something that most TMS-vs-neurofeedback discussions miss entirely, and it genuinely surprised me when I first encountered it in the literature.

There's a small but growing body of research on combining TMS and neurofeedback in the same treatment protocol. And the logic, once you hear it, is almost obvious.

TMS is good at creating an initial shift in brain activity. It can "kick-start" an underactive circuit, like the left DLPFC in depression, forcing it into a more active state. But TMS doesn't teach the brain to maintain that state on its own. It's like jump-starting a car: it gets the engine running, but if the battery is fundamentally weak, the engine stalls again once you disconnect the cables.

Neurofeedback is good at teaching the brain to maintain a state. Once the brain can see its own activity and receive rewards for producing the desired pattern, it gradually learns to stay there. But neurofeedback can be slow to get started, especially if the target circuit is severely underactive. It's hard to train a brainwave pattern that the brain is barely producing in the first place.

So what if you used TMS to get the circuit firing, and then used neurofeedback to teach the brain to keep it firing?

A 2021 pilot study in Brain Stimulation did exactly this for depression patients. Participants received a course of TMS to the left DLPFC, followed immediately by neurofeedback training targeting frontal alpha asymmetry. The combined group showed greater improvement than either treatment alone, and the improvements were more durable at 3-month follow-up.

The researchers' interpretation: TMS opens a window of enhanced neural plasticity in the targeted region. Neurofeedback, applied during that window, helps the brain consolidate the new pattern into a stable, self-maintained state.

This is like the difference between having someone physically position your hands on a piano (TMS) versus learning to play the notes yourself (neurofeedback). The physical positioning can show you where your fingers need to go. But the skill only becomes yours when you learn to do it independently.

The Accessibility Question (This Is the Part That Changes Everything)

Let's step back from the science for a moment and talk about something practical.

If you're reading this article, what can you actually do with this information? What's available to you right now?

TMS is available in clinical settings. You need a referral, usually from a psychiatrist. You need a diagnosis that matches an FDA-cleared indication (depression, OCD, or a handful of other conditions). You need to show that you've tried and failed other treatments first, because insurance typically requires proof of treatment resistance. You need to live near a clinic that offers TMS, and you need to be able to show up every weekday for 4-6 weeks. If you don't have insurance coverage, you need $6,000-$15,000.

These barriers are appropriate for what TMS is: a powerful medical intervention that carries real (if small) risks and requires clinical expertise. TMS should be administered by professionals. Nobody is arguing otherwise.

But here's what that means in practice: TMS is inaccessible to the vast majority of people who might benefit from changing their brain activity. Not because of the science, but because of the logistics.

Neurofeedback has traditionally faced similar barriers. Clinical neurofeedback means finding a practitioner, scheduling appointments, and paying $100-200 per session for 20-40 sessions. That's better than TMS, but it still puts it out of reach for most people.

What's changed is consumer EEG technology.

A device like the Neurosity Crown puts 8 channels of EEG data, sampled at 256Hz across frontal, central, and parietal-occipital regions, into a form factor you can wear at your desk. It provides real-time focus scores and calm scores, which are computed from the exact brainwave features (theta/beta ratios, alpha power, SMR activity) that neurofeedback protocols target. And with open SDKs in JavaScript and Python, developers can build custom neurofeedback protocols tailored to specific training goals.

This doesn't replace clinical neurofeedback for serious conditions. It doesn't replace TMS for treatment-resistant depression. But it opens up a category that didn't exist before: self-directed brain training that you can do at home, on your schedule, for the one-time cost of a device.

Think about what happened when heart rate monitors moved from clinical settings to wristbands. People didn't stop going to cardiologists. But millions of people who would never have visited a cardiologist started paying attention to their cardiovascular health. The technology met them where they were.

That's what's happening with neurofeedback right now.

So Which One Should You Choose?

The honest answer depends entirely on what problem you're trying to solve.

If you have treatment-resistant depression and you've tried medications that haven't worked, TMS has the strongest evidence and the clearest clinical pathway. Talk to a psychiatrist about whether you're a candidate. The SAINT protocol in particular has produced results that were hard to imagine even five years ago. This is the kind of problem TMS was designed to solve.

If you have ADHD and want a non-medication approach, neurofeedback has Level 1 evidence and decades of clinical data. Start with a qualified neurofeedback practitioner who can do a baseline assessment and design a personalized protocol. Consumer EEG devices can supplement clinical training for home practice.

If you're a healthy person who wants to improve focus, reduce anxiety, or understand your own brain better, neurofeedback is the clear winner on every practical dimension. It's accessible, it's safe, the skills transfer to daily life, and the effects persist. A consumer EEG device gives you the core tool you need to start.

If you're fascinated by the idea of combining both approaches, you're not alone. The research on TMS-primed neurofeedback is early but compelling. If you have access to a clinician who offers both, ask about combined protocols.

One thing to be clear about: neither TMS nor neurofeedback is a magic bullet. The brain is the most complex structure we know of, and changing its activity patterns, whether through external stimulation or self-directed training, takes time, consistency, and realistic expectations. Anyone who tells you otherwise is selling something.

This guide is for informational purposes only and does not constitute medical advice. TMS is a medical procedure that should only be administered by qualified healthcare providers. If you are experiencing depression, anxiety, or other mental health conditions, please consult a licensed mental health professional.

The Deeper Question: Done to You, or Done by You?

Here's what stays with me about the TMS-versus-neurofeedback comparison, and it goes beyond the clinical data.

These two approaches represent two fundamentally different philosophies about the relationship between a person and their brain.

TMS says: your brain is broken, and we can fix it from the outside. We know which circuit is underactive. We have a tool that can stimulate it. Sit still and let us work. This is powerful, and for people in genuine neurological distress, it's the right philosophy. Sometimes you need an expert to intervene.

Neurofeedback says: your brain already has the capacity to produce the patterns associated with focus, calm, and well-regulated emotion. It just can't see what it's doing. Give it a mirror, and it will figure out the rest. This is a slower philosophy, and a more demanding one. It asks something of the person. It requires engagement, patience, and practice.

But it produces something TMS cannot: a skill.

When a TMS course ends, the patient leaves with altered neural excitability that will fade with time. When a neurofeedback course ends, the person leaves with a brain that has learned something new about how to regulate itself. The altered excitability from TMS requires maintenance sessions. The skill from neurofeedback becomes part of the brain's repertoire.

We're living through an extraordinary moment in neuroscience. For the first time, ordinary people can measure their own brain activity in real time, at home, without clinical supervision. The same brainwave patterns that neurofeedback clinicians have been training for decades are now visible to anyone with a consumer EEG device and a sense of curiosity.

The magnetic coil is an impressive tool. But the most interesting brain-computer interface might be the one where the brain learns to be its own therapist.