fb-pixel Skip to main content
The Medical Issue

He lost his arm in an accident. A new surgery and a bionic prosthetic are giving him back unprecedented control.

A team of Boston surgeons and scientists has developed a new approach to amputation. “My dream as a scientist is that a person with an arm amputation could play a Beethoven piece at normal speeds and dexterity,” says MIT’s Hugh Herr.

photo illustration by C.J. Burton/for the Boston Globe

IT’S MARCH 2022 and Bradley Burkhard is sitting in an MIT lab, doing his best to follow instructions. “Move your index finger,” says technician Mikey Fernandez, and a finger moves dutifully up and down. “What about the other digits?” he asks, and other fingers curl, a bit more awkwardly. “The thumb?” Fernandez asks. There’s only the barest of perceptible movement. “Yeah, the thumb’s not really doing a whole lot,” the 32-year-old Burkhard sighs, slumped back in an office chair. He’s been at this now for three days, and clearly he’s getting tired.

The fact that Burkhard can move any fingers is practically a miracle. The hand he controls is not his own, but a robotic prosthesis clamped to a lab bench 3 feet away. A tangle of 16 white wires extends back to Burkhard’s residual arm, which ends just above the elbow. The aluminum and rubber prosthesis looks like an android arm from a science fiction movie, and indeed it is called the LUKE arm after the Star Wars hero who famously lost his hand. The next-generation artificial limb, created by Segway inventor Dean Kamen and his team, allows for a finely articulated range of motion. But the real miracle isn’t that arm, it’s Burkhard’s own — and the first-of-its-kind surgery that allows him to control the prosthesis with finely tuned electrical signals from his residual muscles.


As Burkhard’s muscles flex under electrodes connected to those 16 wires, lines of computer code scroll past on a monitor. Fernandez and other scientists at the MIT Media Lab will use that output to calibrate the prosthesis to Burkhard’s motions, in a way they hope will eventually give him an unprecedented amount of control. Ultimately, they plan to attach the artificial limb to Burkhard’s own, and allow him to use it seamlessly, exactly how his arm worked before an ATV accident three years ago.

“Doing amputations kind of sucks,” says Dr. Matthew Carty, a reconstructive plastic surgeon at Brigham and Women’s Hospital. “It has been regarded for thousands of years as a failure — like, I can no longer help this patient by trying to save their limb, so we just gotta cut it off. But we do ourselves and our patients a disservice by thinking about it that way.” This procedure is a new way of thinking.


Carty helped to develop the new surgery to create a so-called agonist-antagonist myoneural interface, or AMI, which reconnects muscles to powerfully amplify electrical signals sent along the nerves. Over time those signals can let a patient superimpose the lingering sensation of a lost limb, often called a phantom limb, over an artificial surrogate. “They can map their phantom limb over the prosthesis,” Carty says, coming to “feel” the prosthesis as if it were part of their biological body.

Carty has performed the AMI procedure as part of several dozen leg amputations in recent years, but Burkhard became only the second person, after a military veteran, to have the operation done on his arm. He’s already showing remarkable progress, working with the MIT Media Lab team led by Hugh Herr, a professor of biomechatronics who came up with the concept for AMI surgery and helped see it to fruition.

“My dream as a scientist is that a person with an arm amputation could play a Beethoven piece at normal speeds and dexterity — and for legs, that a person could dance ballet,” says Herr, who lost both of his own legs in an accident when he was a teenager. “These are the ultimate artistic expressions of physicality that we have in humanity. The field is knocking on the door of that being possible.”


Dr. Matthew Carty (left) examines Burkhard at Brigham and Women’s Faulkner Hospital.Aram Boghosian for The Boston Globe

The night Burkhard’s life changed in April 2019, he was zooming around California’s Malibu Hills with friends in Polaris ATVs, basically souped-up golf carts with roll cages. “We’d been up there hundreds of times, so I didn’t really think much of it,” says Burkhard, recounting his story in a conference room at the MIT lab. An actor and filmmaker, Burkhard is dressed in a pair of green overalls and a T-shirt, with a Carhartt baseball cap pulled down low. His hair is pushed back behind his ears, and he sports a scruffy beard and small hoop earrings.

At the time of the accident, Burkhard was living in Santa Monica, California, working as a bartender and trying to break into the film business. Growing up in Andover, he played basketball and soccer in high school, where he also discovered filmmaking and made a documentary about a young basketball player who had been hit by a drive-by bullet and paralyzed from the waist down. He continued filmmaking while pursuing a sociology degree at Pace University in Manhattan, and left for Los Angeles as soon as he graduated, writing scripts and standing in long lines at casting calls.


On that night in Malibu, Burkhard’s Polaris got stuck in soft ground. As he rocked the vehicle to get it free, it slipped and began tumbling down the hill, flinging Burkhard from the cab. That was really close, he thought as he lay on the ground. As he stood up, he realized that the roll cage had sliced through his left forearm.

“My arm was just hanging there,” he recalls, “like everything was severed except for two nerves.” One of his friends quickly used his belt to make a tourniquet that probably saved Burkhard’s life. At the time, however, he was sure it was the end, accepting his fate even as his friend panicked. “I was like, I’m gonna die,” he recalls. “This is how my story ends.”

He did some breathing exercises he had used to relax before auditions, and after a few minutes began to think he might survive. A life flight helicopter airlifted him to UCLA Medical Center, and he implored his buddies not to call his parents back in Boston (the friends ended up calling anyway). By then the adrenaline had worn off and an excruciating pain had set in. Amazingly, the surgeons were able to reattach his arm. For the next few weeks, he lay in a hospital bed in constant pain while doctors performed one surgery after another. At one point, they attached leeches to help circulate the blood to his arm. (“They would fall off once they were full, and a couple of them got stepped on by people,” Burkhard recalls.) But every day, he felt his fingers less and less — his arm was slowly dying. Ultimately Burkhard gave the medical center’s doctors the go-ahead to amputate. “Once they cut it off,” he says, “my pain went from like an eight to a four.”


Andover native Bradley Burkhard lost his arm in a 2019 accident.guerin blask/for the Boston Globe

When he left the hospital, Burkhard felt optimistic. Money from disability insurance meant he didn’t have to return to work right away, and he spent some of his time traveling. “I basically took a gap year at age 30,” he says. “I had a lot of fun.” He also had time to adjust to having only one arm, learning to cook and dress himself with some difficulty. “I can tie my shoes with one hand, but it takes me like 10 minutes per shoe,” he says. Brushing teeth means clenching the toothbrush in his mouth while he spread toothpaste on the bristles.

Burkhard does have an annoying phantom-limb syndrome, where it feels like he has a hand perched just beneath his elbow, fingers stuck in a claw. “I feel like I’ve got my hand stuck in a Pringles [canister],” he says. But all in all, he feels lucky. “It was close to being a lot worse — I could have been dead,” he says, noting that he’s also fortunate not to have lost his dominant right arm. “If you are going to lose a limb, you probably want to lose your off arm.”

As the disability money ran out, though, Burkhard began to run into his limitations. Typing was slow, and bartending with only one hand seemed impossible. He got by working as a production assistant on movie sets, and later working at a hotel in New York City. Making work more challenging, his arm healed with only a thin flap of skin at the end, which had a tendency to split and bleed when he bumped it. “I was just constantly having to carry Band-Aids,” Burkhard says.

His parents, however, learned of a surgeon at Walter Reed National Military Medical Center in Bethesda, Maryland, named Jason Souza. He told them about a new kind of amputation surgery that was showing promise with military veterans who had lost legs in combat, giving them renewed feeling and control over their artificial limbs. The surgery had been done successfully for several years for legs, but never for arms.

Dr. Souza referred Burkhard to Matthew Carty, a colleague at Brigham and Women’s, who had pioneered the procedure. Carty told Burkhard he could give him a new amputation above his elbow. It would mean losing even more of his arm, but it could potentially lead to improved function of a prosthetic arm.

Burkhard was wary about going back into the operating room, but a new procedure that promised more than he’d been able to get from a traditional one intrigued him. “My main thing was, I didn’t want to ever have to do [a surgery] again — so why not get the newer one?”

Matthew Carty, who is 50, had been working for years to restore limbs for people who had been injured in wars, as well as motorcycle accidents and other misfortunes — even transplanting hands, arms, and legs from organ donors. And for years, he’d been frustrated that surgeons seemed to treat amputation as the end of a journey rather than the beginning. “Amputation cases are always assigned to the most junior person,” Carty says, “because it’s not considered intellectually interesting or gratifying.”

While technologies for transplants and artificial limbs had been getting more sophisticated over the decades, they were still limited by a kind of amputation surgery that hasn’t changed much. “Your options are, to a great extent, informed by what you do at the time of amputation,” Carty says, speaking in his office at Brigham and Women’s, where his shelves are crammed with diplomas and awards. “It’s the staging ground for what happens afterward.”

For a while, Carty had been thinking that there must be a better way to improve quality of life for patients, but it wasn’t until the Boston Marathon bombing in 2013 that he finally decided he had to do something about it. “We took care of a lot of patients who had lower extremity injuries,” says Carty, who sports a head of premature white hair and a friendly bedside manner. “That really crystalized it. I said, Man, we really gotta think about a new way to do this.”

A colleague introduced Carty to Hugh Herr at the MIT Media Lab, who had been working for years to create better limb replacements, and the two met to imagine what a better outcome might be.

Herr had also been frustrated by the limitations of traditional amputation surgery. Arm prostheses fall into two main categories: Body-powered limbs use a harness and cables to control the limb with muscles elsewhere, such as the chest or shoulders — an awkward solution at best. Myoelectric limbs, in contrast, use electrodes pressed against the skin in a socket to pick up faint electrical signals when the user tenses a muscle; those signals control motors in the artificial limb. Though they’ve been used for some 40 years, such artificial limbs are often slow and can feel unnatural to patients who use them, only allowing a few movements, and requiring users to switch between individual motions such as opening and closing fingers and rotating the wrist. Crucially, they also don’t allow for any feedback that can tell users what their arm or hand is doing without looking at it.

Artificial limb manufacturers have compensated with more advanced robotics and machine learning to interpret intentions and more fluidly combine movements, but such modifications have so far proved unreliable. That leads many people after an arm amputation to just give up on using an artificial limb entirely.

Herr knows how alienating a prosthesis can be. When he was 17, he got lost while climbing Mount Washington in New Hampshire, and was forced to spend the night in freezing temperatures. Frostbite claimed both of his legs.

In the 40 years that followed, Herr dedicated his life to creating new artificial limbs for himself and others, earning a master’s degree in mechanical engineering at MIT and a PhD in biophysics from Harvard. That has put him in the unique position of being able to design, test, and use new prototypes. “I am acutely aware of the impact of bad design on persons trying to use technology — and also of good design,” he says. The first time he tried a robotic ankle himself, it was a disempowering experience. “It felt like I was being walked, not doing the walking,” Herr says. “It’s like if you could never drive a car, but you are always in the back seat being driven.”

Hugh Herr at MIT, with bionic legs of his own design.matthew Septimus

When he met Carty, he’d been thinking about how to restore that sense of agency, following humans’ natural biology rather than making them more of a machine. Knowing that the movement of muscles profoundly amplifies nerve impulses, Herr wondered if there was a way during amputation surgery to knit pairs of muscles back together to retain their movements — and boost the signals coming from them. As the bicep contracts, for example, the tricep extends as if attached to a pulley — it’s what anatomists call an agonist-antagonist relationship. With a typical amputation, however, those muscles are severed and sewn into position, where they are allowed to “scar in place.”

Herr asked Carty if it would be possible to retain that muscle motion. Sure, Carty said, from a surgical perspective such a connection should be easy to create. The next day, he came back with sketches to demonstrate. Together they refined the concept into the idea of an agonist-antagonist myoneural interface, or AMI. (“Myoneural,” from the Greek, means relating to both muscles and nerves.) Theoretically at least, they could create a stronger signal to more naturally control a prosthesis in real time, while simultaneously giving feedback to the brain on how the limb was moving.

Herr, who also codirects the K. Lisa Yang Center for Bionics at MIT, is soft-spoken and serious, where Carty is outgoing and confident. But the two hit it off in a way that they say a clinical surgeon and bioengineer rarely do. “It’s so rare surgeons talk to me — it’s almost without precedent,” Herr says. “It’s a lesson in how people with different backgrounds from different institutions can work together.”

After the Marathon bombing, Steven and Audrey Epstein Reny donated money to Brigham and Women’s to create the Gillian Reny Stepping Strong Center for Trauma Innovation, named for their daughter, whose legs were injured in the bombing. Now director of strategy and innovation for the center, Carty used part of the Reny grant to help conduct experiments in Herr’s lab at MIT.

The beauty of the AMI procedure (pronounced “Amy”) is as much the electrical impulses sent from the muscles as those sent to them. “When I hold my hand behind my head and wiggle my fingers,” Carty says, demonstrating, “I know where my hand is — in fact, I could probably catch a ball.” That’s because nerves within the muscle transmit specific information to the brain that tells it where our limbs are in space. “The technical name of that is proprioception,” he says, and it’s a key to the AMI surgery. Carty and Herr hypothesized that by recombining the muscles’ natural pulleys, patients would for the first time be able to feel the position of an artificial limb just as they would their biological limb, and that in turn would give them greater control over its movement.

In 2016, Carty conducted the first AMI surgery on Jim Ewing, a Maine native who’d lost his leg in a climbing accident — giving the lower-limb procedure its name, an Ewing amputation. Since then, he’s done 34 lower limb operations at both Brigham and Women’s and Walter Reed, and the procedure has produced exactly the results they’d hoped for. During a review of 22 leg patients last year, Carty and Herr found improved perception and control compared with patients undergoing traditional amputation. In fact, many Ewing amputees mapped their phantom leg sensation completely onto their prosthesis, feeling it like their own. “One of our lower limb patients was hiking and said he went through a creek and felt a sense of water on his prosthetic foot,” Carty says. “[I]t was remarkable when he talked about it, because he had such a developed sense of his phantom.”

Meanwhile, Herr’s lab has continued to develop the neural interface to better interpret the signals coming from the muscular nerves, and use them to operate the motors within prosthetic legs. “We’re demonstrating the first fully brain-controlled leg ever,” Herr says.

Though legs can be challenging to design given the stress placed upon them by body weight, that’s nothing compared with the complicated machinery of the arm. While current upper limb prostheses can allow a person to control arm movements in sequence — first the elbow, then the wrist, then the hand — the human arm doesn’t work that way; even a simple act like reaching out and grabbing a cup of coffee takes tremendous coordination between different muscles in the arm and hand, all moving in concert.

“How we interact with our hands in the world is different than how we interact with our feet,” Carty says. “So the demands for a meaningful prosthesis are ten times higher.” In July 2020, an Iraq war veteran underwent AMI surgery for an arm at Walter Reed, with an amputation below the elbow that proved successful. A few months later in October, Burkhard underwent his own surgery at Brigham and Women’s. Burkhard was the perfect candidate for an AMI, since Carty could perform a completely new amputation above the elbow, cutting above the original, traditional amputation. Losing the elbow joint, however, would make control of a prosthetic arm even more difficult.

Burkhard didn’t feel any nervousness going into the hospital for the AMI procedure, but just as he was being wheeled into the operating room, he felt a twinge of anxiety. “I had a similar sensation once when I went skydiving with my friends,” he says. “I wasn’t nervous until actually jumping out and free-falling.”

At MIT, Burkhard practices controlling the LUKE arm prosthesis (in foreground) through electrodes attached to his residual arm.Jimmy Day/MIT Media Lab

In September 2021, nearly a year after his AMI surgery — and five months before his experiments at MIT with the LUKE arm — Burkhard traveled back to Boston for a follow-up at the Brigham. He had finally gotten a prosthetic arm that summer, two years after his first amputation at UCLA. Putting it on “was definitely an emotional experience,” he says. He finally felt like he had two arms again.

But he quickly found the prosthetic was often more trouble than it was worth. The electronics inside made it bulky and heavy, and it was a few inches longer than his biological arm. Worst of all, it had the tendency to slip off his residual arm, and it took a long time to work it back into the socket, say, at the grocery store. “I don’t really want to try and find some place to put it on in front of a bunch of strangers,” he says.

In an exam room at the Brigham he stands before a table with a series of tasks laid out — needle and thread, tape dispenser, scissors. “I’m probably not going to be super-awesome at this,” he warns occupational therapist Kerstin Palm, admitting that he’s really only used the prosthesis seriously for a day. Nevertheless, she encourages him to try. In one task in which he needs to flip over six index cards, he does it within seconds with his biological hand. But he struggles with his prosthesis, his fingers awkwardly rolling over a card, trying to get a grip. He fiddles with an iPhone app that can change the position of the finger grip, and the hand eerily spins around in a 360-degree maneuver before he can get it under control.

Finally, he grasps a card, and pulls it to the edge of the table. Just as it seems it might fall off, he flips it with his forefinger. “I think [the prosthesis] might have slipped off, because the nerves aren’t lining up,” Burkhard says, pulling off the cuff to reveal a tan, muscular upper arm. He sprays it with a special lubricant before refitting the black plastic sheath. After that, he slides another card to the edge of the table, prosthetic hand grasping at the air for minutes before he’s able to finally position his thumb and forefinger to clutch it and flip it over. He flips the next two cards more quickly.

In just about an hour of practice, Burkhard visibly improves at other tasks, dropping bottle caps into a canister with aplomb. “I know that must have been frustrating, but you have unbelievable precision,” Palm says. “The fact that you can control the device at all after one day is encouraging.”

Carty was able to create “two and a half” AMIs for Burkhard — one to control the elbow, and one to control all his fingers together; the “half” is a backup of another workable muscle that he might learn to control for another purpose. Today, almost a year and a half after the surgery, Carty is pleased with Burkhard’s progress from a medical perspective. Truly getting to the point where Burkhard feels comfortable with the prosthetic arm and starts to feel like it is his own, however, will take another layer of work. That’s where the experiments at Herr’s MIT lab come in.

While not specifically designed for the AMI procedure, the LUKE arm is designed to allow for a much finer grade of movement than Burkhard’s current prosthesis, including the potential for multiple simultaneous motions of the wrist, elbow, and fingers. It was created by Dean Kamen, the New Hampshire inventor best known for creating the Segway two-wheeled transporter, which he originally designed to help people with disabilities. The LUKE arm — an acronym for Life Under Kinetic Evolution — is a continuation of that work.

MIT’s version has a programmable interface that Herr and his research assistants can use to customize the response to the signals Burkhard’s AMI sends out.

Already in the short time he’s been in the lab, Burkhard has surprised the researchers with the amount of control he’s been able to show. Carty figured the AMI for Burkhard’s fingers would only be able to control all of his digits at once — but Burkhard has been able to show individual finger control, and even slightly move his thumb. Even so, Burkhard doesn’t have any illusions that he will have a truly functioning arm without more work. “I always knew it wasn’t going to be anything in the next five years,” he says. “But now I feel like in my lifetime I’ll actually have another arm back.”

Carty has now performed AMI amputations on four upper-extremity patients. Encouragingly, he’s found that other surgeons around the world are starting to perform their own versions of the surgery, which might soon become the standard in the field. “These are patients who, until now, didn’t really have many other options and faced living a life that was less than what they hoped to achieve,” he says. “I have no existential angst about my day-to-day contribution to society.”

Recently, Carty even performed something he initially thought impossible: a “revision” surgery, in which he was able to go back into a patient who had already had a traditional amputation and create an AMI to restore the functionality of the muscles. He has a grant to further explore such procedures, potentially opening up eligibility for the new technique to a wider range of patients.

Meanwhile, Herr is pursuing a new procedure now undergoing Food and Drug Administration review in which 3-millimeter magnetic spheres would be implanted into the muscles during surgery. A prosthesis could then use magnets to measure the minute distance between spheres within, allowing it to even more precisely convert the signals into perception and control. “I’m hoping we can really find the right combination of soft tissue surgery and magnetomicrometry so a person could again play Beethoven,” he says.

For his part, Burkhard looks forward to one day being able to play basketball, among other activities he used to take for granted. “It would be nice to be able to tie a tie,” he says. At least as far as acting goes, he now has something unique to set him apart in casting calls. He’s in the process of making a movie about a sports gambler who gets in trouble after a bad bet, and is working on a screenplay for a slasher film.

He’d like to get to the point of having a working arm so he doesn’t have to field sometimes prying questions about what happened to him. “There’s definitely days where it’s been a long day and people would ask and I’m like, ‘I don’t really want to talk about it.’” Even so, he says, it’s not for himself but for future generations of amputation patients that he is most excited to continue the experiments. “I feel bad for the 6-year-old kid who loses his arm — you have to have two limbs when you are playing outside and all that stuff,” he says. “If I can help get this technology to a point where it’s really high-functioning, then maybe that 6-year-old can have an arm and live a normal life.”

Michael Blanding is a frequent contributor to the Globe Magazine and the author of the book “North by Shakespeare.” Send comments to magazine@globe.com.