Imagine reaching deep into the brain—past the skull, past the delicate outer layers—without a single incision. That’s the promise driving one of the most intriguing experiments in neuroscience right now, and it hinges on a tool most people associate with pregnancy scans: ultrasound.
At West Virginia University’s Rockefeller Neuroscience Institute, neurosurgeon Ali Rezai is testing whether low-intensity focused ultrasound can quiet the cravings that drive substance use disorders. Rezai spent a 30-year career pioneering deep brain stimulation, performing more than 3,000 surgeries that involve drilling into the skull and implanting electrodes. His first such operation, back in 1996, took a grueling 16 hours. Ultrasound offers a radically different path.
How does sound reach that deep? Modern focused ultrasound machines fire dozens of beams from different angles through an intact skull. Each beam originates from piezoelectric crystals that vibrate rapidly when hit with voltage, generating high-frequency sound waves. Individually, no single beam carries enough energy to harm the bone or tissue it passes through. But converge them at one point—with millimeter precision, controlled by a mouse click—and they reach maximum intensity exactly where you want them.
The idea isn’t new. In the 1950s, brothers William and Francis Fry built a two-room apparatus that focused ultrasound like a magnifying glass concentrates sunlight, burning pinprick lesions to relieve Parkinson’s tremors. But their method required removing part of the skull, and once the drug levodopa arrived in the 1970s, the surgery fell out of favor.
What sets Rezai’s work apart is targeting. Other noninvasive techniques, such as transcranial magnetic stimulation, only reach the brain’s outer layers. Rezai is aiming for the nucleus accumbens—a pea-sized structure buried in the reward circuitry that magnetic stimulation simply can’t touch.
- The procedure takes about 20 minutes, and patients go home the same day.
- No drilling, no incisions, no probing of brain tissue.
- Rezai’s team uses an FDA-approved high-intensity machine with the power turned way down, starting low and increasing gradually.
In 2021, with FDA permission, Rezai launched an exploratory safety trial. The early signs were striking. “The cravings went down live on the table,” Rezai says. During sessions, participants view drug-related cue images designed to provoke craving, rating their intensity on a 0-to-10 scale via hand controllers while a helmet-shaped transducer delivers the waves. Because hair can scatter ultrasound and blunt the effect, patients shave their heads beforehand.
The context makes the stakes obvious. Nearly 45 million Americans meet the criteria for a substance use disorder, and 2025 saw close to 70,000 overdose deaths. West Virginia carries the highest overdose rate in the country—more than double the national average—which is precisely why Rezai relocated there despite colleagues warning it was “academic suicide.”
This remains experimental research, not an approved therapy, and trials include placebo groups plus real risks—dizziness, headaches, even brain hemorrhage. But the notion of rewiring the brain’s reward system without a scalpel is one of the boldest bets in modern medicine.
