The power of thought
Implant translates brain signals from two paralyzed patients into control of a robotic arm, fueling hope of independence
Cathy Hutchinson imagined picking up her coffee from the table. She thought hard about bringing the red bottle toward her lips and taking a drink, without any assistance. Then, for the first time since a stroke left her arms and legs paralyzed 15 years earlier, she did it.
A blue robotic arm, guided by an experimental brain implant that “read’’ Hutchinson’s thoughts, grasped the bottle and carried it toward her. By picturing her own immobile right arm and hand moving, she navigated the robot arm to the right position, tipped the bottle toward her lips, and took a long, satisfied sip through a straw.
A smile spread across her face. Researchers who watched in rapt silence in April 2011 burst out in applause.
On Wednesday, the team, led by scientists from the Providence VA Medical Center, Brown University, and Massachusetts General Hospital revealed that two paralyzed patients - Hutchinson, 59, of East Taunton, and a man named Robert - had used their minds to control a robotic hand, successfully touching and grasping objects in repeated trials.
Researchers said this was the first published report of using such technology to manipulate a robot arm to reach for and grab objects.
“I had feelings of hope and a great sense of independence, when drinking from a cup,’’ Hutchinson wrote this week in response to reporters’ questions. She answered by choosing letterswith her eyes to spell out words, because the stroke also left her unable to speak. “This was a long term goal I had and I was elated.’’
The feat, reported in the journal Nature, was part of an ongoing clinical trial of a pill-size device covered in electrodes called the BrainGate. It is implanted in the brain to record the activity of brain cells triggered by thinking. Those electrical signals are decoded by computer algorithms, and translated into the movement of a cursor on a screen or a robotic arm through the air.
The ultimate goal is restoring mobility and independence to people who have been paralyzed through spinal cord injuries, disease, or strokes.
“The smile on her face is something I and our whole research team will never forget,’’ said Dr. Leigh Hochberg, a researcher at the VA and Brown and a critical care neurologist at Mass. General. “The progress both our participants have demonstrated is very encouraging, though there’s undoubtedly still more work to do.’’
It will be years before a device could be commercially available, and it is possible that day may never come. Even at this early point, the road has been filled with unexpected turns.
The proof that neural signals could be translated into controlled, 3-D movement in a real-world environment is the product of years of research into basic neuroscience, engineering, and robotics. It’s the result of millions of dollars of private investment and federal funds. A Foxborough company, Cyberkinetics Inc., founded to commercialize the technology, went out of business three years ago during the economic bust, a setback that did not stop the clinical trial, which is now being administered by Mass. General.
The new work built on previous accomplishments by the same team of researchers, which demonstrated in 2006 that the device, implanted in the brain of a research subject, could be used to move a cursor around on a screen, open e-mail, and operate a television. Other research groups are also working on technology to bridge the devastating gap that can occur when paralysis or disease leaves people’s minds intact but their bodies unresponsive. Those alternate approaches include an external cap fitted with sensors that read the brain’s electrical activity and electrodes implanted under the skull.
In the BrainGate experiment, the small device was implanted near an area of the brain involved in voluntary muscle movements. The two research subjects first watched the robot arm mounted on an adjacent platform perform a series of movements while imagining themselves moving their own arms in the same way. The chips in their brains measured the pattern of electrical activity from dozens of neurons, detecting what signals meant move right or left, or squeeze an object.
Then the researchers gave control of the arm to Hutchinson and Robert. With their thoughts, both participants used a robotic arm to try to repeatedly touch and grasp a succession of pink foam balls affixed to cones. In a separate trial, the researchers gave Hutchinson the opportunity to serve herself the drink. She succeeded in four out of six attempts over 8 1/2 minutes.
Outside scientists said the experiments were an exciting advance, and added it was significant that both patients had their injuries for some time because it shows the necessary brain circuits remained functional. Hutchinson had been implanted with the BrainGate device for five years, also suggesting that the technology could work over a long span. Robert had suffered a stroke 5 1/2 years before the study and had the implant for five months.
“You could imagine someone who had an amputation or a spinal cord injury five or 10 years before would not be eligible . . . but these patients show the competency still remains,’’ said Story Landis, director of the National Institute of Neurological Disorders and Stroke, which began funding the foundational research by Brown University neuroscientist John Donoghue in 1986.
The current study was supported by the National Institutes of Health and the Department of Veterans Affairs, and Landis said she thought the technology might have important applications for veterans wounded in Iraq or Afghanistan, with the possibility of one day providing a precise way to control prosthetic limbs.
Still, challenges remain. The biggest problem now facing the field, which includes technologies that use neural implants and those that detect electrical activity using external sensors, is the need for consistency, said Dr. Jonathan Wolpaw, chief of the laboratory of neural injury and repair at the Wadsworth Center at the New York State Department of Health.
“The dominating feature for the field is that all these approaches face the same, very big issue of reliability,’’ Wolpaw said in an e-mail. “None of these methods is anywhere near as reliable and consistent as it would need to be to be of use in everyday life.’’
For example, Hutchinson successfully touched the pink foam balls within the allotted time about half the time with one robotic arm and a little over two-thirds of the time with another type of robot arm. Over all the trials, she successfully grasped the objects less than half the time. Robert touched the foam balls in 96 percent of the trials, and grasped them successfully 62 percent of the time.
Still, for Hutchinson, the thrill of seeing her thoughts seamlessly translated into movement made her feel “ecstatic.’’ She said she dreams of one day having the ability to control her legs and hopes to one day be able to resume her hobbies of gardening and cooking.