For the first time, a small group of patients with amputations below the knee were able to control the movements of their prosthetic legs through neural signals—rather than relying on programmed cycles for all or part of a motion—and resume walking with a natural gait. The achievement required a specialized amputation surgery combined with a non-invasive surface electrode connection to a robotic prosthetic lower leg. A study describing the technologies was published today in the journal Nature Medicine.
“What happens then is quite miraculous. The patients that have this neural interface are able to walk at normal speeds; and up and down steps and slopes; and maneuver obstacles really without thinking about it. It’s natural. It’s involuntary,” said co-author Hugh Herr, who develops bionic prosthetics at the MIT Media Lab. “Even though their limb is made of titanium and silicone—all these various electromechanical components—the limb feels natural and it moves naturally, even without conscious thought.”
The approach relies on surgery at the amputation site to create what the researchers call an agonist-antagonist myoneural Interface, or AMI. The procedure involves connecting pairs of muscles (in the case of below-the-knee amputation, two pairs), as well as the introduction of proprietary synthetic elements.
The interface creates a two-way connection between body and machine. Muscle-sensing electrodes send signals to a small computer in the prosthetic limb that interprets them as angles and forces for joints at the ankle and ball of the foot. It also sends information back about the position of the artificial leg, restoring a sense of where the limb is in space, also known as proprioception.
Video 1
www.youtube.com
“The particular mode of control is far beyond what anybody else has come up with,” said Daniel Ferris, a neuromechanical engineer at the University of Florida; Ferris was not involved in the study, but has worked on neural interfaces for controlling lower limb prostheses. “It’s a really novel idea that they’ve built on over the last eight years that’s showing really positive outcomes for better bionic lower legs.” The latest publication is notable for a larger participant pool than previous studies, with seven treatment patients and seven control patients with amputations and typical prosthetic legs.
To test the bionic legs, patients were asked to walk on level ground at different speeds; up and down slopes and stairs; and to maneuver around obstacles. The AMI users had a more natural gait, more closely resembling movement by someone using a natural limb. More naturalistic motion can improve freedom of movement, particularly over challenging terrain, but in other studies researchers have also noted reduced energetic costs, reduced stress on the body, and even social benefits for some amputees.
Co-author Hyungeun Song, a postdoctoral researcher at MIT, says the group was surprised by the efficiency of the bionic setup. The prosthetic interface sent just 18 percent of the typical amount of information that’s sent from a limb to the spine, yet it was enough to allow patients to walk with what was considered a normal gait.
Next Steps for the Bionic Leg
AMI amputations have now become the standard at Brigham and Women’s Hospital in Massachusetts, where co-author Matthew Carty works. And because of patient benefits in terms of pain and ease of using even passive (or non-robotic) prosthetics, this technique—or something similar—could spread well beyond the current research setting. To date, roughly 60 people worldwide have received AMI surgery above or below either an elbow or knee.
In principle, Herr said, someone with a previously amputated limb, such as himself, could undergo AMI rehabilitation, and he is strongly considering the procedure. More than 2 million Americans are currently living with a lost limb, according to the Amputee Coalition, and nearly 200,000 lower legs are amputated each year in the United States.
On the robotics side, there are already commercial leg prosthetics that could be made compatible with the neural interface. The area in greatest need of development is the connection between amputation site and prosthesis. Herr says commercialization of that interface might be around five years away.
Herr says his long-term goal is neural integration and embodiment: the sense that a prosthetic is part of the body, rather than a tool. The new study “is a critical step forward—pun intended.”
