Erin McNulty had been missing for weeks when her mother, Linda, sat down on a chair in her living room, exhausted. Linda had put in her usual seven-day workweek at the antiques shop she runs near Burlington, Vermont. She’d spent her free evenings driving around, trying to track down her daughter. Erin, 45 at the time, had been using methamphetamine for years.

Her substance use started in high school—first alcohol, then marijuana, and eventually heroin. Erin’s brother used heroin, too. When Linda found out, she started driving her kids to a methadone clinic three hours away in Massachusetts. The methadone helped, but it made Erin feel tired all the time, so she started using cocaine to stay awake. There were stretches of sobriety—she had her daughter during one of them, in 2008. The family took trips to the Great Escape waterpark in New York and Hampton Beach in New Hampshire. There were also several overdoses and attempts at rehab. Eventually, Erin switched to suboxone and stopped using heroin. But when a friend introduced her to meth, Linda says, “Erin was gone.”

Linda turned on the TV and was flipping through the channels when a 60 Minutes segment caught her attention. It was early 2024, and the show focused on a procedure that might help people with substance use disorders. Linda immediately thought of her daughter. The procedure involved beaming ultrasound through the skull to treat the brain. Researchers at West Virginia University were testing it on people with Alzheimer’s disease and addiction.

The neurosurgeon behind the procedure, Ali Rezai, was a pioneer in the field of deep brain stimulation, which involves cutting into the skull to implant electrodes that can reach neurons deep in the brain. He was excited by the ability of ultrasound—the imaging tool best known for observing fetal development during pregnancy—to reach into the same brain structures without breaking the skin. There would be no drilling into the skull, no poking or prodding the brain’s delicate tissue. The ultrasound could be delivered in 20 minutes, and patients could go home the same day.

Linda couldn’t believe what she was seeing on TV. She yelled out to her teenage granddaughter, Erin’s child, to come watch. “Look at this doctor,” she said. “They’re fixing people!” It didn’t matter that the trial was happening more than 600 miles away. She was going to get Erin to West Virginia. But first she had to find her.

Rezai, who was born in Iran, grew up in Los Angeles, and completed his neurosurgical training in New York, never expected to end up in West Virginia. “My colleagues said, ‘You’re committing academic suicide,’” he says. But Rezai saw an opportunity to pursue new treatments at the epicenter of the US addiction crisis. Almost 45 million Americans meet the criteria for a substance use disorder, and although drug overdose deaths in the US have been declining for the past three years, there were nearly 70,000 such deaths in 2025. West Virginia has the highest rate of overdose deaths in the country by population, more than double the national average.

Over the course of his 30-year career, Rezai has carried out more than 3,000 deep brain stimulation surgeries. (He stopped counting after that.) He’s used the approach to treat Parkinson’s, severe obsessive-compulsive disorder, traumatic brain injury, and more recently, opioid addiction. His first DBS operation, during his medical residency at New York University in 1996, took a painstaking 16 hours. Dozens of computers were crammed into a basement room to record from the patient’s brain. Rezai says they had to operate in the basement because of all the beeping, clicking, and whirring. Today, these devices can be implanted in a fraction of the time with just a few computers—but brain surgery is still brain surgery.

The first attempts to harness ultrasound to treat the brain required brain surgery too. In the 1950s, brothers William and Francis Fry built a system that could focus ultrasound beams into a single point, similar to how a magnifying glass can focus scattered sunlight. The whole apparatus took up two rooms, with the controls upstairs and the ultrasound probes coming through the ceiling. Ultrasound waves generate heat at high intensities, and the brothers used their system to burn pinprick-sized lesions in the brains of animals without damaging surrounding tissue. They did the same in patients with Parkinson’s disease to destroy tiny portions of brain tissue responsible for abnormal movement. It relieved people’s tremors, but the procedure never caught on because it required removing part of the skull. By the 1970s, the drug levodopa became commercially available to treat Parkinson’s, making surgery an unnecessary risk.

Modern focused ultrasound machines can emit dozens of beams from different angles through an intact skull. When an electric voltage hits their piezoelectric crystals, they vibrate rapidly, producing high-frequency sound waves. With the click of a mouse, the waves can be targeted to a spot in the body with millimeter precision. On their own, individual beams don’t produce enough energy to affect the bone or tissue they pass through. But when they converge into a single point at the targeted depth, they reach their maximum intensity.

High-intensity ultrasound, like what the Fry brothers used, is already approved to treat a handful of conditions, including uterine fibroids, certain kinds of prostate cancer, and symptoms of Parkinson’s that can’t be treated with drugs. But ultrasound can also be delivered in low intensities to alter neural activity in the brain.

The ability to get deep into the brain without holes or incisions is why Rezai and others are so excited about ultrasound. Other types of noninvasive stimulation, such as transcranial magnetic stimulation, only breach the outer layers of the brain. They can’t penetrate the nucleus accumbens, the pea-sized part of the brain Rezai is targeting for addiction. Rezai wanted to see if focused ultrasound could.

In 2021, with permission from the Food and Drug Administration, Rezai and his team at WVU launched an exploratory trial to test the safety of focused ultrasound in volunteers with addiction. They used an FDA-approved machine designed to emit high-intensity waves but turned the power way down, starting off with a low dose and gradually increasing it.

Rezai had no idea how long the effects of a one-time ultrasound procedure might last. Almost immediately, the approach showed promise. “The cravings went down live on the table,” he says. The volunteers started to report feeling, well, different.

With the TV segment on Rezai occupying her thoughts, Linda intensified her efforts to find Erin. A few days later, Linda found her in a hotel in downtown Burlington. Erin had been staying with an ex-boyfriend. “She was in pretty rough condition,” Linda says.

She brought Erin home to their house in the country and told her about the study. Erin watched the video over and over. She felt hopeful, but she was also scared. She had experienced anxiety and depression for much of her life, and drugs were a way to cope. She felt her most normal when she was taking drugs, which is what kept her using.

“It absolutely took over,” Erin says. “I couldn’t get out of it.” The drive to get meth overrode everything else in her life, including being a mother and putting a roof over her head. She wanted to get better, but she couldn’t imagine not doing meth anymore.

Linda was convinced that Rezai’s study was her daughter’s way out. To be considered, Erin needed to be in active treatment within the WVU Medicine system. Linda planned to fly with her daughter and granddaughter to Pittsburgh, the closest major airport to Morgantown, and drive the rest of the way, about an hour and a half. But it took three tries to get Erin on the plane. Their first flight was early in the morning, and they didn’t make it to the airport on time. During the second attempt, they were waiting to board the plane when Erin said she had to use the bathroom. Instead, she walked out of the airport. The third time, Erin’s daughter held her hand during boarding and didn’t let go.

They made it to Morgantown, and in April 2024 Erin checked into the Center for Hope and Healing, a residential treatment facility affiliated with WVU, which was recruiting for Rezai’s study. When her mother and daughter left, Erin was miserable.

“I wanted this, and I knew it was the answer, but still, there was a part of me that wasn't comfortable with even the thought of being without my drug of choice,” she says. “I just thought my life was going to be over.” She was only at Hope and Healing for a week when her cravings got so intense that she left. She eventually came back to the treatment center.

From their home in Vermont, Linda and her granddaughter made calls and sent emails to the WVU team, pleading with them to accept Erin into the ultrasound trial. “My mom wants this procedure more than anything,” Erin’s daughter wrote in an email to Rezai in September 2024. “She cannot live without this.”

Eventually Erin was stable enough to join the study. By October, she was lying in an MRI machine anticipating a blast of ultrasound. Rezai’s team explained the risks involved in the procedure—including potential dizziness, headaches, and brain hemorrhage—and that Erin might land in the placebo group and not receive a real treatment at all. Erin agreed to all of it. Hair can scatter ultrasound waves, potentially dulling the treatment effect, so she needed to shave her head as part of the protocol. She hesitated at first but ultimately decided it would be worth it.

As Erin lay on the MRI table, Rezai and his team showed her pictures of drugs she’d used in the past—heroin, fentanyl, methamphetamine, and benzos, chosen by the research team to trigger her specific cravings. “We want to put them in the maximum pathological state,” Rezai explains. “That's why we show these pictures. It's called priming.” With every image, he asked Erin to rate her craving level for that substance on a scale from 0 to 10.

A frame held Erin’s head in place under a helmet-shaped transducer as researchers went through the motions of delivering the ultrasound. She didn’t know it, but she had been assigned to receive the sham—a fake treatment meant to serve as a control so that researchers could evaluate how well the ultrasound was working.

Rezai and the team checked her cravings again. They were still high. Afterward, she stayed positive. She wanted to believe that she had received the procedure she’d been waiting so long for. But the cravings came back, and a few weeks later, she was using meth again.

She was desperate to get better, and went back to Hope and Healing. Three months later she received a call. She learned she had received the sham treatment and was being invited back to get the real thing. It was a happy moment for Erin. She was still hopeful that the procedure would help her.

In January 2025, she was back under the MRI machine, head shaved again. Erin was ready. The machine’s motorized table slid her inside the scanner, and Rezai showed her the images again. Ultrasound waves can warm up the skull, so the team monitored temperature changes in her brain tissue on the MRI.

This time, researchers beamed the ultrasound to her nucleus accumbens, roughly 8 centimeters from the top of the head. Highly addictive drugs like meth and opioids flood the nucleus accumbens with the feel-good neurotransmitter dopamine. Over time, the brain gets used to these surges and needs more of the addictive substance to get the same effects. “This part of the brain involved in reward gets dysregulated and hypersensitized. It's always front and center, seeking the dopamine high,” Rezai says. The ultrasound would attempt to reset her nucleus accumbens and break the connection her brain had made to meth.

While she was still in the machine, Erin’s cravings dropped off. She was no longer connecting with the images of drugs.

The procedure was all over in 20 minutes. Erin moved to a sober living home, and in the following months, Rezai and his team periodically tracked her craving levels through a mobile app and took urine samples to test for drug use. She saw people using drugs on the bus and on the street, but it didn’t faze her. It was like a switch in her brain had been turned off.

Rezai and his team have now performed focused ultrasound on 47 people with substance use disorders as part of ongoing trials. In the first 20 participants, who all knew they were receiving ultrasound, the team observed a more than 90 percent reduction in self-reported cravings and an 80 percent reduction in positive urine tests. Data from the randomized controlled trial, which Erin was a part of, is still being analyzed, but Rezai says they are seeing consistent results. The WVU protocol is now being adopted by other medical centers, with trial sites in Baltimore, Florida, and New York, as well as internationally. Rezai’s group is also testing the approach for binge eating disorder, and he wants to adapt it for use in dogs with anxiety.

Although still highly experimental, focused ultrasound’s ability to alter brain activity has captured the interest of Silicon Valley tech founders. Sam Altman’s Merge Labs, Coinbase cofounder Fred Ehrsam’s Nudge, Mary Lou Jepsen’s Openwater, and Sanmai Technologies, backed by LinkedIn founder Reid Hoffman, are all aiming to miniaturize large ultrasound machines into wearable devices to treat a range of psychiatric and mood disorders without drugs or brain surgery—a potentially huge market. The Focused Ultrasound Foundation estimates that there are over 100 device manufacturers and hundreds of ongoing clinical trials for different indications.

Rezai dreams of expanding access to focused ultrasound across West Virginia, performed by trained technicians rather than highly specialized neurosurgeons like himself. WVU is now testing a portable device on wheels that doesn’t require a huge MRI machine, and researchers there are building an in-house device that resembles a salon hair dryer.

Before focused ultrasound can become ubiquitous, researchers have a lot of questions to answer. One is what, exactly, focused ultrasound is doing to the brain. The leading theory is that ultrasound might be acting on certain cellular gatekeepers known as ion channels that allow sodium, potassium, and calcium into cells. Altering that flow changes how neurons fire, which can directly affect emotions and behavior.

There's also the question of, how much ultrasound is too much? Last year, about a month after Erin’s procedure, a 44-year-old volunteer in the WVU trial suffered a brain injury in the MRI machine. While researchers were increasing the ultrasound’s intensity, the man suddenly became unresponsive. The procedure was immediately stopped. He regained consciousness 24 hours later and, over the next few weeks, became alert and attentive again but experienced periodic confusion. He still struggles with memory issues. Scans showed that the man suffered tiny hemorrhages in and around his nucleus accumbens.

The incident brought the trial to a halt for eight months while Rezai and the team consulted with their data safety monitoring board, the FDA, outside researchers, and the manufacturer of the ultrasound device, Israel-based Insightec. Rezai sent a letter describing the injury to a scientific journal, prompting a flurry of responses. The authors said Rezai’s letter left out key safety details such as the intensity level of the ultrasound, leaving the field in the dark about how the injury happened and what should be done to help prevent serious side effects in the future.

Greg Fonzo, assistant professor of psychiatry and behavioral sciences at the University of Texas at Austin, describes the incident as a “black mark” on the field. Ultrasound can sometimes cause headaches or tingling at low intensities, but a brain injury is highly concerning, he says. Fonzo thinks the problem was that the machine used in the WVU trial was too powerful for what they were using it for.

To make the system safer, Rezai’s group worked with Insightec to add a mechanism that drops the power back down once it reaches a certain threshold. It would be like a car reducing speed automatically if the driver went over the speed limit. Another added safety feature turns the ultrasound off entirely if it’s too high for too long. He described the safety mechanisms in a follow-up letter to the journal.

This is one of the biggest gaps in research on focused ultrasound. There are no standard treatment parameters. And with dozens of companies in the space, devices operate at different frequencies and intensities. Many studies are early-stage, and investigators are still trying to find the most effective, and safest, dosage. Fonzo, for instance, is using a small tabletop device to study focused ultrasound for depression, anxiety, and post-traumatic stress disorder. In a trial he ran at UT Austin, participants received low-intensity ultrasound once a day, five days a week over the course of three weeks—similar to how transcranial magnetic stimulation is given.

Rezai says that regimen won’t work for patients with substance use disorder, since there's a risk of patients dropping out of treatment programs early. Requiring too many sessions increases the likelihood that patients won’t come back to finish treatment, so Rezai thinks a one-and-done approach is ideal for treating addiction.

Kim Butts Pauly, a professor of radiology and electrical engineering at Stanford University who wrote one of the responses to Rezai’s letter, says the WVU team is operating at a midrange level where less is known about the effects. She said it’s possible that their approach is actually causing structural changes in the brain. Pauly is working on a research proposal to develop a system for monitoring skull and brain temperature to avoid serious side effects.

Rezai says ultrasound is not a magic cure for addiction. Patients also need behavioral therapy, as well as support during recovery. And a person’s environment can make the difference in whether they maintain recovery. A few study participants who relapsed after getting ultrasound reported using drugs again, not because they were craving them but because they were available and someone offered them. It’s also unknown whether patients will need repeated ultrasound treatments.

Erin stayed in West Virginia after her procedure. She got a job at Goodwill processing donations and rented an apartment. She became absorbed in decorating her bedroom to look like it’s “out of a magazine”—a knack she acquired from working at the family antiques business. She took a course to become a peer support specialist for other people struggling with substance abuse.

Erin felt like she was thriving. She liked living in a college town, taking public transportation, and working out regularly. But she was ready to go home. This past May, she moved back to Vermont with the help of her mom and daughter.

Her hair has long since grown back, and Linda gushes about how Erin’s outfits are always so put together. She’s working at the antiques shop and considering a job as an evening caregiver. She’s living with her mother and daughter again, and her free time is spent chasing their five rambunctious golden doodles. Erin still has an exercise routine, too—Linda often finds her on the Gazelle machine on their deck.

She says that she doesn’t think about drugs anymore, and that she doesn’t have the desire or cravings she once did. She feels like someone who has never used them. “I never imagined it could be this good,” Erin says. Her daughter is relieved to have her mom back. She doesn’t have to worry anymore about where she is.

The three of them don’t talk about Erin’s past drug use. They’re focused on the present. Linda is convinced that the ultrasound procedure removed the cravings from her daughter’s brain for good—a “reboot,” as Rezai calls it.

Now, grandmother, daughter, and granddaughter are planning a beach trip to Maine. They can do that now, plan for the future.

Let us know what you think about this article. Submit a letter to the editor at [email protected].