CAMBRIDGE, Mass. — For hours on end last year, Massachusetts Institute of Technology engineers ran the brain-controlled robotic limb through its paces, testing its capabilities on a series of patients and fine-tuning it like a pit crew preparing a race car for the Indy 500.
They had patients flex the prosthetic foot: toe up, toe down. Toe in, toe out. Walk up a flight of stairs, then back down. But it was after the day’s experiments, when patient Jim Ewing was seated and chatting with the team, that they made their most provocative observation: He fidgeted, pivoting the motorized ankle, unconsciously.
It wasn’t much, just a wiggle now and then, but it provided powerful evidence that the new robotic foot had become a seamless part of Ewing’s body in a way that has never been accomplished before with a prosthetic limb, the scientists reported Wednesday.
“A standard amputee, when they’re wearing their leg, doesn’t do any of that,” Dr. Matthew Carty, a Brigham and Women’s Hospital surgeon and co-author of the paper, told STAT. “But Jim, when he was sitting there talking to us, was fidgeting his bionic foot like it was his biological foot. And he wasn’t thinking about it. It was just him being him.”
The key to making the bionic like the biological was combining a surgical advance with a technological one, the researchers say. Ewing — who mangled his left foot when he fell about 50 feet from a cliff he was scaling in the Cayman Islands — was the first person to undergo an entirely new kind of amputation, pioneered by Carty and MIT professor Hugh Herr. The engineers, meanwhile, developed a prosthetic foot that would enable two-way communication, with signals traveling from Ewing’s brain to his residual lower leg and into the bionic limb, and then back again.
Herr, himself a rock climber who lost both his legs to frostbite as a teen, describes his goal as nothing short of eliminating disability. That’s still a ways off. But the research demonstrates the potential of the new system to help people with amputated legs catch themselves when they step off curbs they don’t see, or hike across uneven terrain without stumbling, said Tyler Clites, who just earned his Ph.D. from the Harvard-MIT Program in Health Sciences and Technology and was lead author of the paper published in Science Translational Medicine.
Ewing, referred to as “Subject A” in the paper, was the only patient tested who underwent the new procedure. Clites compared how he did with the robotic foot to the performance of four people who had traditional amputations. In contrast to Ewing, who described the prosthesis as “his leg,” the others reported “a distinct lack of ownership of the prosthesis, or emotion associated with controlling it,” Clites wrote.
Ewing, 54, exhibited more subtle control over the carbon-fiber-and-metal foot, which weighs about 2 kilograms — roughly the same as his natural foot did. He also automatically raised and lowered his toes while climbing or descending stairs, a reflexive behavior seen in people with intact legs but never before with a prosthetic one. “This is below the level of his consciousness,” Clites said. “It’s just happening, naturally.”
Before concluding that the new surgery is superior to the standard amputation, however, the findings must be replicated in a larger clinical trial, which is already underway.
STAT has been following the experiment for the last 15 months, collecting footage for an upcoming documentary, called “Augmented,” about Herr and his team’s ambitious project to create advanced robotic prosthetics that become one with patients.
Most crucially, the MIT and Brigham researchers were able to restore Ewing’s sense of proprioception, the ability to know — without looking — where your limbs are in space, how fast they’re moving, and with what force. It’s what allows you to touch your fingers together with your eyes closed, or to calibrate just how hard to press on the gas pedal. In short, “it is fundamental to all human movement,” Clites said in an interview.
Paul Marasco, head of the Laboratory for Bionic Integration at the Cleveland Clinic, called the paper “cool,” saying “proprioception has been a really tough nut to crack” for developers of robotic limbs. “It’s a sense that we don’t even know we have,” he said, “that we take completely for granted. But without it, the functionality of prosthetics is really reduced.”
In people with intact limbs, proprioception depends on the interplay between opposing muscles: When your biceps contract and move your elbow, for example, the triceps stretches, and sensors in those muscles and attached tendons signal your brain that your arm has bent. This continuous feedback allows the brain to finely control movement, but in the traditional amputation procedure, these connections are severed.
The new “Ewing amputation” recreates them, by sewing together pairs of muscles and tendons that used to connect to the ankle and the subtalar joint in the foot. When one muscle contracts, the other stretches, restoring the sensation that the joint is moving. Electrodes attached to the patient’s skin record the muscle movements in his residual limb and send signals to the prosthetic foot that cause it to move the way he feels like it’s moving.
At the same time, the foot sends back electrical impulses to the residual limb about the force being exerted, enabling the patient to adjust how hard he is pushing.
“What’s unique about Jim versus others is that when Jim thinks about moving his phantom limb, he feels as if the phantom limb is moving the way he wants it to move,” Clites said.
Ewing could feel the sensation as soon as the robot foot was attached and tuned up. “I right away started using it as if it were my own foot,” he told STAT, without having to retrain his brain how to get the desired movements.
“Oh, wow, there’s something there,” he recalled thinking. “It’s responding. It felt kind of like my foot had returned.”
“I’m not somebody who’s really prone to dramatic emotions,” he added, “but later on when I was driving home, I really felt a strong urge to be connected to it again. Like, ‘I want to have that thing on again and feel like I have my foot back.’”
Ewing, an engineer from Falmouth, Maine, was so focused on the “monumental” decision about whether to have his leg amputated at all that undergoing the novel procedure was almost an afterthought. He couldn’t be sure the surgery would work, that it would eliminate the excruciating pain that was causing him to take increasing amounts of pain pills. And what if it made the pain even worse?
“It’s not reversible, you can’t put it back on,” he said. “That’s the scary part.”
But on July 19, 2016, a year and a half after his fall, he went ahead with the amputation: “I knew that I couldn’t continue through life with what I had,” he said.
Now two years later, his pain has largely subsided, though he said he still feels “nerve noise,” a “constant little bit of tingling in the ghost foot.” It doesn’t slow him down. Since the operation, he’s gone skiing and scuba diving. He runs and hikes with a carbon fiber prosthetic, and, pretty much every weekend, he goes climbing. “There’s really no activity that I have to say, ‘Oh, I can’t do that because of my injured limb.’”
The standard amputation surgery remains mired in the 19th century: It hasn’t changed much since the days of the Civil War, when anesthesia was still rudimentary and what mattered most was speed and keeping patients alive.
Meanwhile, prosthetic limbs have improved dramatically in the last decade or so, driven largely by the need to help the thousands of American service members who have returned from the wars in Afghanistan and Iraq with shattered legs. The U.S. Army helped fund the research described in the new paper, along with the MIT Media Lab Consortia, Google, and the Gillian Reny Stepping Strong Center for Trauma Innovation, which was created by the family of a Boston Marathon bombing survivor.
“The technologies have matured to the point where we can start having the conversation about how amputations need to change” to take full advantage of the features of advanced prosthetics, said the Cleveland Clinic’s Marasco. “That’s a really exciting piece of this [paper].”
Marasco recently reported restoring the sensation of hand movement in three patients with upper limb amputations. By vibrating muscles in the patients’ residual arm while their prosthetic hand moved, Marasco’s team tricked the brain into thinking it was feeling the hand move. Patients had better control of their grip, and, similar to Ewing, reported that their bionic limb felt more like their own.
Clites said his team’s approach is the only one that relies on muscles and nerves doing what they do naturally. (He has filed for a patent with Herr and Carty, who already hold a conceptual patent on the procedure.)
“Everyone else who is working in this space is trying to build a robotic system that will function in a residual limb that frankly is broken,” Clites said. “Instead of doing that, we’re going in and we’re reengineering the physical body of the patient so that it is optimized to interact with the prosthetic limb.”
One afternoon early last year, after watching Ewing successfully climb stairs with the robotic leg, Herr sat in an MIT conference room and reflected on what he’d seen. “When we design and build hammers and we pick them up and we drive nails, it’s a tool,” he said. “It’s separate from our bodies. It’s something we use. But it’s not an integral part of self. We’re now entering a new era of human-technology interaction.”
He said he yearns for that sensation of ownership, what he calls “embodiment,” and he admitted feeling “green with envy” for Ewing. “But the time will come,” he added.
Indeed, last year, Herr’s lab published a paper describing how the new approach to amputation could be applied to people like Herr who’ve already undergone the standard procedure.
Carty is already planning to start performing an above-the-knee version of the Ewing amputation, with the first patient scheduled for surgery in June and two more likely in the fall.
These patients will be part of the ongoing larger clinical trial, which is funded with $3 million from the Department of Defense. The trial will ultimately include 16 to 20 patients, Carty said. It was supposed to take four years, but recruitment has gone much faster than anticipated, driven by word of mouth. Carty has so far done nine of the amputations.
Two years ago, Ewing limped into Brigham and Women’s Faulkner Hospital for the first one. Earlier this month, he returned, this time striding confidently down the hall on a bionic leg, its small electric motor whirring with each step. He was there to visit a 20-year-old woman who was about to undergo the amputation procedure bearing his name.
As he turned the corner into the preoperative area, patient No. 9 was waiting eagerly. “Hi, Jim,” she said.
“Did you hear my parts,” he asked.
Then they embraced.