The Promise of Brain-Machine Interfaces
WHY YOU SHOULD CARE
Soon, bionic limbs will restore mobility. Soon after, minds will move objects. Before you know it, you’ll be sending that text or email with just your thoughts.
By Melissa Pandika
When 32-year-old Zac Vawter lost his right leg in a motorcycle wreck in 2009, the simple act of climbing stairs became a tricky balancing act. To avoid falling, he had to go up each step with his left leg first, hoisting his stiff, prosthetic leg behind him.
Now the Seattle father of two can not only climb stairs, he can scale a ramp and kick a soccer ball with ease — thanks to a bionic leg that can produce a full range of ambulatory movements by reading his intention to move from a bundle of nerves just above his knee.
“He can blend in and do the same things everyone else … is doing,” said bioengineer Annie Simon, a member of the Rehabilitation Center of Chicago team that designed the limb. “It’s amazing.”
Once the stuff of Star Wars movies and Isaac Asimov novels, such brain communication systems — known as brain-machine interfaces or BMIs — have become a promising possibility. These systems could allow disabled individuals to control not only prosthetic limbs but also wheelchairs and computers. In the more distant future, BMIs may even mean that people can learn to play a musical instrument in their sleep or communicate using only their thoughts.
These systems could allow disabled individuals to control not only prosthetic limbs but also wheelchairs and computers.
The human brain controls movement by sending electrical signals to the muscles. But that flow of information is disrupted in individuals with brain or spinal cord damage. BMIs fix those broken connections, translating the intention to move into instructions that guide an external device, like a prosthetic limb.
Vawter’s leg had been severed at the knee, meaning that the nerves that originally sent electrical signals to his ankle stopped at his damaged knee muscle. But those nerves eventually recovered and began carrying electrical impulses again, so neurosurgeons simply rerouted them to his healthy hamstring muscle. Vawter then spent hours imagining himself making specific movements with his missing limb, causing the redirected nerves to generate distinct patterns of contractions in his hamstring, which were detected by sensors on his skin. The researchers programmed, or “taught,” the 10-pound robotic leg to recognize each contraction pattern as a specific motor command.
Most prosthetics require users to toggle switches or cables to transition from one movement to another. But Vawter only needs to imagine such movements, the research team reported in the New England Journal of Medicine last month.
Most prosthetics require users to toggle switches or cables, but Vawter only needs to imagine movements…
Vawter is using the first thought-controlled bionic leg, but scientists have also designed arms that can maneuver like their human equivalent. Last December, a woman paralyzed for 10 years from the neck down used a robotic arm to grab and bite into a chocolate bar. Since the nerves in her arms didn’t function at all, University of Pittsburgh neurosurgeons implanted electrode sensors in her brain, where they decoded signals from her neurons into motor commands that they sent to the robotic arm.
Meanwhile, scientists at the UC-San Francisco and UC-Berkeley Center for Neural Engineering and Prostheses, or CNEP, are developing a prosthetic sleeve — called an “exoskeleton” by researchers — that, with the help of implanted electrodes, would allow people to reanimate their paralyzed arms. The team hopes to launch a clinical trial in January, according to CNEP neuroscientist Karunesh Ganguly.
Other types of BMIs that don’t signal to prosthetic limbs could open the door to more mainstream use. CNEP scientists are working on a system that would enable people to control a computer cursor with just their thoughts — making it possible to send emails and text messages from a smartphone without even reaching for the device.
Some BCIs may even go a step further, enabling communication without words or sound. In a pilot study published in August, one University of Washington neuroscientist sent a brainwave through the Internet that caused another researcher to move his right index finger. Scientists at Boston University and ATR Computational Neuroscience Laboratories in Japan believe a similar process could one day teach people how to play the piano or throw a curveball by modifying their brain activity patterns to match those of someone who has already mastered such skills.
To those envisioning Matrix-like mind control, “We’re still very far from decoding thoughts,” said Amy Orsborn, a bioengineer at UC-Berkeley and UCSF. Besides, “These devices are not for mind control. They’re for providing new means of control for sensory processes.”
Scientists still face several roadblocks to developing truly natural-feeling BMIs. It can take weeks to learn how to use a BMI and then the device needs to be “retaught” to recognize different brain signaling patterns with each use. Researchers are working to shorten the learning times so that users are good to go almost as soon as they strap on the device.
Others propose embedding the brain with thousands of tiny sensors, or “neural dust” particles.
What’s more, brain implants have been shown to detect electrical signals for approximately five years. Since the brain is floating in a sea of cerebrospinal fluid, it moves around, but implants remain fixed to the skull, so the neurons used to “teach” the system gradually shift away from the electrodes, causing their signal to weaken until they’re barely detectable. Some bioengineers are designing more flexible electrodes that bend with the brain as it moves. Others propose embedding the brain with thousands of tiny sensors, or “neural dust” particles, which are more likely to stay in contact with the neurons used to calibrate the system.
Another hurdle is that unlike natural limbs, BMIs don’t deliver sensory feedback to the brain. But Duke University researchers say it’s possible, reporting in September that monkeys felt a ball brushing their arm when they saw an avatar arm being touched by a virtual ball, offering hope that they can engineer prosthetics that feel natural and integrated with the rest of the body. Meanwhile, Simon and her colleagues are designing their bionic leg so that it can change how stiff it is depending on the ground surface — allowing more give on grass versus concrete, for example.
Participate: The University of Pittsburgh Department of Physical Medicine and Rehabilitation, which led the trial for the robotic arm, is always looking for clinical trial participants. Contact Debbie Harrington at email@example.com or (412) 383-1355 for more information.
While Vawter can’t bring his bionic leg home just yet, Simon and her colleagues hope to roll it out for home trials in about five years. Devices that rely on brain implants will take longer, possibly a decade or more, experts say. But who knows? If January’s clinical trial for CNEP’s prosthetic sleeve proves successful, things “can move really quickly,” Ganguly said.
- Melissa Pandika Contact Melissa Pandika