What Is Focused Ultrasound?

The Surgeon's Impossible Dream

For most of the history of neurosurgery, there's been a cruel tradeoff. The deeper a brain structure sits, the more tissue you have to cut through to reach it.

The thalamus, for instance. It's a walnut-sized relay station buried near the center of your brain, responsible for filtering sensory information before it reaches your cortex. When the thalamus misfires, you get tremors so severe you can't hold a fork. For decades, the only way to fix a misfiring thalamus was to open the skull, navigate past cortical tissue that was working perfectly fine, and destroy or stimulate the tiny cluster of neurons causing the problem.

Neurosurgeons got remarkably good at this. But the fundamental problem never went away: to help one brain region, you had to physically violate others. Every millimeter of the surgical path carried risk. Infection. Bleeding. Damage to fibers that happened to be in the way.

Now imagine someone told you there was a way to reach that same thalamic target, with millimeter precision, without breaking the skin. No incision. No drill. No electrode threaded through healthy tissue. Just a patient lying in an MRI machine while focused beams of sound converge on a point deep inside their brain and do the job from outside.

You'd think it was science fiction. But it's been FDA-approved since 2016.

This is focused ultrasound. And it is, without exaggeration, one of the strangest and most promising technologies in all of neuroscience.

Sound as a Scalpel: The Physics

Here's the core idea, and once you understand it, everything else about focused ultrasound clicks into place.

You know how a magnifying glass works. It takes sunlight, which is arriving from many angles, and bends all those light rays so they converge on a single point. At that focal point, the energy is concentrated enough to burn paper. But if you hold your hand an inch above or below the focal point, you feel nothing unusual. The light is there, but it's too spread out to do anything.

Focused ultrasound does the same thing with sound waves instead of light.

A device called a transducer array (shaped like a helmet and containing over 1,000 individual ultrasound elements) sits against the patient's head. Each element emits an ultrasound wave. Individually, these waves are too weak to affect tissue. They pass through skin, bone, and brain matter harmlessly, just like the diffuse sunlight above and below the magnifying glass's focal point.

But the computer controlling the array adjusts the timing of each element so that all 1,024 waves arrive at the same target at the same instant. At that convergence point, typically 3 to 5 millimeters wide, the acoustic energy adds up. And depending on how much energy you deliver, very different things happen.

The beauty of all three regimes is the same: the intervening tissue, everything between the transducer on the outside and the focal point on the inside, experiences only low-level, harmless ultrasound. The damage (or the modulation, or the barrier opening) happens only where the waves converge.

Why Not Just Use TMS? (Or Any Other Stimulation Method)

If you've been following the brain stimulation world, you might be thinking: we already have transcranial magnetic stimulation, transcranial direct current stimulation, transcranial alternating current stimulation, and a whole alphabet soup of non-invasive techniques. Why do we need another one?

The answer is depth.

Every other non-invasive brain stimulation technique shares the same fundamental limitation. They can only effectively reach the cortex, the wrinkly outer layer of the brain that sits just beneath the skull. TMS generates magnetic pulses that induce electrical currents in neurons, but the induced field drops off so sharply with distance that it can only meaningfully stimulate tissue within the first 2 to 3 centimeters. tDCS and tACS push electrical currents through the scalp, but those currents spread diffusely and weaken as they travel deeper. None of them can precisely target a structure at the center of the brain.

This is not a small limitation. Many of the most important brain structures, the ones implicated in movement disorders, depression, addiction, and memory, are deep. The thalamus, the hippocampus, the basal ganglia, the amygdala, the nucleus accumbens. All of them sit well below the cortical surface.

Look at that table for a second. Before focused ultrasound, if you wanted to reach a deep brain target with any precision, you had exactly one option: surgery. Deep brain stimulation required drilling holes in the skull and threading electrodes through brain tissue. It worked remarkably well for the patients who needed it, but the barriers to entry were enormous. General anesthesia. Multi-hour surgery. Six-figure costs. Risk of infection and hemorrhage. A permanent implant that needed battery replacements.

Focused ultrasound broke that paradigm. For the first time, clinicians could reach deep brain targets non-invasively with millimeter-level precision. It didn't replace DBS for all cases (some patients still need the continuous stimulation that an implanted electrode provides), but it opened a door that had been locked for a century.

The Blood-Brain Barrier Problem (And How FUS Cracks It Open)

Here is, in my opinion, the single most mind-bending application of focused ultrasound. And the one most likely to change medicine in the next decade.

Your brain has a bouncer. It's called the blood-brain barrier, and it is spectacularly good at its job.

The blood-brain barrier is a layer of tightly packed endothelial cells lining every capillary in your brain. These cells are connected by structures called tight junctions, and those junctions are so impermeable that they block roughly 98% of all small-molecule drugs and nearly 100% of large-molecule drugs (antibodies, gene therapies, most chemotherapy agents) from crossing from the bloodstream into brain tissue.

This is great for keeping toxins and pathogens out of your brain. It's terrible when you actually need to deliver medicine there.

Think about it this way. We have chemotherapy drugs that can kill glioblastoma cells in a petri dish. We have antibodies that can clear amyloid plaques associated with Alzheimer's disease. We have gene therapies that could potentially fix the root cause of Huntington's disease. But getting any of these therapies into the brain in therapeutic concentrations? The blood-brain barrier says no. And that "no" has been one of the most frustrating bottlenecks in all of neuroscience and oncology.

This is where focused ultrasound, combined with a simple trick involving microbubbles, changes everything.

Here's how it works. The patient receives an intravenous injection of microbubbles, tiny gas-filled spheres about the diameter of a red blood cell. These microbubbles circulate harmlessly through the bloodstream. Then, focused ultrasound is aimed at the specific brain region where you want the drug to enter.

When ultrasound waves hit the microbubbles inside the brain's capillaries, the bubbles oscillate. They expand and contract rhythmically with each pressure cycle of the sound wave. This oscillation is gentle enough that the bubbles don't pop (that would be too violent), but vigorous enough to mechanically push and pull on the endothelial cells lining the capillary walls. The tight junctions between cells stretch and temporarily loosen.

For about 24 hours, the blood-brain barrier at that exact location becomes permeable. Drugs in the bloodstream can cross into brain tissue. Then the tight junctions reseal, the barrier closes, and the brain goes back to its default state of extreme selectivity.

The precision here is what makes it extraordinary. You're not opening the blood-brain barrier across the entire brain (that would be catastrophically dangerous). You're opening a window the size of a pea, exactly where you need it, for exactly as long as you need it. It's like having a key that only opens one lock in a building with ten thousand doors.

How Researchers Actually Do This (The MRI Partnership)

You might be wondering: if the focal point is only a few millimeters wide and it's buried deep inside the brain, how do you know you're aiming at the right spot?

The answer is MRI. And the partnership between focused ultrasound and magnetic resonance imaging is one of those engineering marriages that feels almost too perfect.

Here's how a typical MRI-guided focused ultrasound procedure works for thermal ablation (the FDA-approved essential tremor treatment):

The patient's head is fixed in a stereotactic frame inside an MRI machine. A hemispheric transducer array filled with degassed water is positioned around the skull. The MRI provides a real-time, three-dimensional map of the brain. The clinical team identifies the target, usually the ventral intermediate nucleus of the thalamus (Vim) for tremor.

Then comes the clever part. Before delivering any therapeutic energy, the system fires a low-energy "test shot" and uses a special MRI sequence called MR thermometry to measure the temperature change at the focal point. This lets the team verify that they're hitting exactly the right spot before committing to the full-power ablation. They can adjust the focal point, account for skull thickness variations, and confirm that no unintended tissue is being heated.

Only when everything checks out do they increase the power and perform the therapeutic sonication. The entire time, MRI is watching the temperature of the tissue in real time. If anything drifts off target, the team can stop instantly.

This is brain surgery with a live map and a pause button. No scalpel in history has had those features.

Neuromodulation: The Quiet Revolution

The FDA-approved applications of focused ultrasound (ablating tremor-causing neurons, opening the blood-brain barrier) tend to grab the headlines. But there's a subtler application that might end up being the most impactful: low-intensity focused ultrasound neuromodulation.

At very low intensities (well below what would heat tissue or open the barrier), focused ultrasound can temporarily modulate neural activity. The exact mechanisms are still being worked out, but the leading theory involves mechanosensitive ion channels. These are protein channels in neural membranes that open or close in response to mechanical force. When an ultrasound pressure wave passes through a neuron, it physically deforms the membrane by nanometers. That deformation is enough to gate these mechanosensitive channels, which can either excite or inhibit the neuron depending on the specific channel type and the ultrasound parameters.

The parameters matter enormously. Pulse frequency, duty cycle, intensity, duration, and sonication timing all influence whether the effect is excitatory or inhibitory. Researchers have shown they can use low-intensity FUS to:

• Suppress activity in the somatosensory cortex, reducing the brain's response to touch
• Enhance motor cortex excitability, making it easier to trigger a muscle contraction
• Modulate the thalamus in humans, shifting sensory perception thresholds
• Alter mood-related circuits by targeting deep limbic structures that TMS can't reach

And here's the part that gets neuroscientists genuinely excited: these effects are temporary and reversible. You can turn a brain region "off" for 30 minutes, study what happens, and then let it come back online. This is an incredibly powerful research tool. It's like having a reversible, non-invasive, spatially precise lesion study on demand.

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

Clinical trials are now exploring low-intensity FUS for conditions that have historically been difficult to treat: treatment-resistant depression (targeting the subcallosal cingulate), OCD (targeting the internal capsule), epilepsy (targeting seizure foci), and chronic pain (targeting the thalamus). The advantage over TMS for all of these is the same: FUS can reach the deep structures that are actually implicated in the disorder.

The Skull Problem (And How They Solved It)

If focused ultrasound sounds too clean, too elegant, too good, here's the catch. And it's a significant one.

The skull is a terrible medium for ultrasound.

Ultrasound waves travel beautifully through soft tissue. They travel beautifully through water. But bone absorbs, scatters, and distorts ultrasound in complicated ways. The skull isn't uniformly thick. It's not uniformly dense. It has air-filled sinuses. Different regions have different ratios of cortical bone to trabecular bone. Every skull is shaped slightly differently.

When ultrasound waves pass through bone, three things happen that you don't want. First, some of the energy gets absorbed, which means less reaches the target and more heats the skull itself (not ideal). Second, the waves slow down in bone compared to soft tissue, which shifts their phase and defocuses the beam. Third, the irregular skull geometry scatters the waves in unpredictable directions.

If you just fired 1,024 ultrasound beams at someone's head without accounting for any of this, you'd get a blurry, shifted, degraded focal point that might miss the target by centimeters. That's not acceptable when you're trying to ablate a 3-millimeter cluster of neurons.

The solution is called aberration correction, and it's one of the most computationally impressive things in medical imaging. Before every procedure, the system takes a CT scan of the patient's skull. A computer model calculates exactly how each of the 1,024 transducer elements' waves will be affected as they pass through that specific patient's skull at that specific angle and thickness. Then it adjusts the timing (phase) and amplitude of each element individually to compensate.

In other words, the system pre-distorts the outgoing waves in exactly the right way so that, after the skull distorts them in the opposite direction, they arrive at the target perfectly focused. It's the acoustic equivalent of adaptive optics in astronomy, where telescopes deform their mirrors to cancel out atmospheric distortion.

This skull correction is what turned focused ultrasound from a theoretically beautiful idea into a clinically viable technology. And it took decades of physics, engineering, and computational modeling to get right.

Where Things Stand: The Research Landscape in 2026

Focused ultrasound sits at a fascinating inflection point. One foot is firmly in approved clinical practice. The other is in the research frontier.

The most closely watched trials in 2026 involve blood-brain barrier opening for Alzheimer's disease. Several groups have shown that repeatedly opening the barrier over amyloid plaques, combined with anti-amyloid antibodies, can reduce plaque burden in targeted regions. The question now is whether plaque reduction in specific regions translates to cognitive improvement. If it does, focused ultrasound could become a cornerstone of Alzheimer's treatment.

The neuromodulation research is earlier-stage but equally exciting. The challenge is reproducibility. Because the effects of low-intensity FUS depend so heavily on exact parameters (frequency, duty cycle, pressure, duration, brain target), different labs using slightly different protocols sometimes get different results. The field is working toward standardized parameter sets, but it's not there yet.

And then there's sonogenetics. This is the one that makes even seasoned neuroscientists do a double take. Researchers have engineered neurons to express mechanosensitive proteins that respond specifically to ultrasound. Then they use focused ultrasound to activate only those engineered neurons, leaving all the others untouched. It's like optogenetics (using light to control neurons), but with sound instead of light. And sound penetrates the skull far better than light does. Sonogenetics is still in animal models, but the implications for understanding brain circuits are enormous.

Measurement and Stimulation: Two Sides of the Same Coin

Here's something that gets lost in the excitement about stimulation technologies. Stimulation without measurement is flying blind.

If you change something in the brain, you need to know what changed. Did the intervention actually shift neural activity in the targeted region? By how much? For how long? Did it affect any other regions? Is the effect consistent across sessions?

This is where EEG comes in. And it's not a minor detail. It's the foundation that makes stimulation research interpretable.

EEG and focused ultrasound are not competing technologies. They're not even in the same category. FUS is a stimulation tool. EEG is a measurement tool. One pushes. The other listens. And the combination of pushing precisely and listening carefully is what separates real neuroscience from guesswork.

Researchers already pair EEG with focused ultrasound in virtually every neuromodulation study. They record baseline brain activity before stimulation, monitor changes during stimulation, and track how quickly the brain returns to baseline after stimulation ends. EEG's millisecond temporal resolution makes it ideal for this, since the neural effects of focused ultrasound unfold over seconds to minutes, and you need something fast enough to capture the dynamics.

The Neurosity Crown fits into this picture as a practical EEG monitoring tool. Its 8 channels at 256Hz, positioned across frontal, central, and parietal-occipital regions, capture the kind of broadband neural oscillation data that's most relevant to assessing the effects of brain stimulation. Its on-device processing means you can track changes in real time. And its open SDKs in JavaScript and Python mean researchers and developers can build custom monitoring applications tailored to their specific stimulation protocol.

You don't need a $50,000 clinical EEG system to track whether your brain's oscillatory patterns shifted after an intervention. You need good sensors, good data, and the ability to look at it in real time.

The "I Had No Idea" Part

Here's something that genuinely surprised me when I first encountered it, and I think it captures just how counterintuitive focused ultrasound is.

In 2023, researchers at the University of Virginia demonstrated that focused ultrasound blood-brain barrier opening could activate the brain's immune system in a way that helped clear Alzheimer's-associated proteins, even without delivering any drug at all.

Just the act of opening the barrier briefly triggered a local immune response. Microglia (the brain's resident immune cells) became activated and started engulfing amyloid plaques near the opening site. The barrier opened, the immune system noticed, and cleanup crews showed up.

This was not the intended effect. The original goal was to use the barrier opening as a delivery mechanism for drugs. The immune activation was a bonus. But in some studies, the barrier opening alone, without any therapeutic payload, produced measurable reductions in amyloid plaque burden.

Think about what that means. The blood-brain barrier, which we've always thought of as a passive wall, might actually be part of the brain's signaling system. Opening it briefly might be telling the brain's immune cells: "Hey, something unusual is happening here. Come check it out." And those immune cells might do useful work once they arrive.

This is still being investigated. Nobody's claiming focused ultrasound alone cures Alzheimer's. But the finding suggests that the blood-brain barrier is more dynamic and more functionally interesting than anyone assumed. It's not just a wall. It might be a communication channel.

What Comes Next

The trajectory of focused ultrasound resembles the early days of laser surgery. When lasers were first proposed as surgical tools, the idea sounded absurd. You're going to shoot a beam of light at someone and call it surgery? But the physics was sound, the engineering caught up, and now laser procedures are so common that we don't even think of them as remarkable.

Focused ultrasound is on a similar path, but with an acoustic wavelength instead of an optical one. The physics has been proven. The engineering works. The FDA has approved the first applications. Clinical trials are expanding into a dozen new indications. The question is no longer whether focused ultrasound works. It's how far it can go.

The most exciting near-term possibility is the development of portable, lower-cost focused ultrasound systems. Current systems (the Insightec Exablate) require a full MRI suite and cost millions. But low-intensity neuromodulation doesn't need the thermal ablation power levels or the real-time thermometry. Several research groups are working on smaller, lighter transducer arrays that could bring FUS neuromodulation into clinic rooms or, eventually, home settings.

If that happens, the combination of focused ultrasound stimulation and real-time EEG monitoring becomes genuinely practical. Stimulate a specific brain circuit. Watch, in real time, how the brain's electrical activity responds. Adjust. Repeat. This is closed-loop neuromodulation, and it's the direction the entire field is heading.

We're not there yet. But the pieces are falling into place faster than most people realize. The skull problem is solved. The targeting is precise. The safety profile for low-intensity applications looks clean. The deep brain structures that matter most for psychiatric and neurological disease are now reachable without surgery.

For the first time, we can both reach the brain's deepest structures and read its electrical responses from the surface. We can push and listen at the same time. And that combination is going to change what's possible.

The brain's fortress has a new door. And we're just starting to learn what's on the other side.