Unlocking the brain
Assistant Professors P. Charles Garrell (left) and Justin Williams implant sensor arrays (inset) on epilepsy patients' brains to study how they learn to communicate via their thoughts.
(View larger image)
f you think about lifting your arm, your brain uses neurons (the wires of the body's intricate information network) to send its electrical information rocketing through your body to your arm, which complies with the appropriate action.
But for people with ALS, high spinal-cord injuries or brain-stem strokes, the neurons that transmit "move, please" requests from the brain to the body
are damaged. So while these patients' brains are intact, the electrical signals that control voluntary motion — everything
from raising an arm or leg to respiration and speech — are locked inside.
Collaborating with researchers at Washington University, the University of Washington and the Wadsworth Center at the New York State Department of Health, Assistant Professors Justin
Williams and P. Charles Garell (also
director of functional neurosurgery at UW Hospital) and BME graduate students
Elizabeth Felton and Adam Wilson are studying a unique combination of mind and machine that may help people with those diseases regain their ability
to communicate — and ultimately, to move.
Currently, says Williams, people with advanced ALS or other conditions communicate
via eye blinks or other nonverbal means — or with the help of speech synthesizers or
rudimentary computer systems that exploit patients' existing communication functions.
In search of a better solution, he and Garell are addressing the problem's root. "We're trying to make a new communication
path by tapping into parts of these people's brains that are still active," he says.
The key is placing tiny neurosensors, which detect neurons' electrical signals, directly on certain parts of the brain. So if one sensor were located near a neuron that controls
arm motion, a patient could think about moving her arm and the sensor would detect the neuron's signal and send the information via wires to a processor to translate it into an action. In the case of an ALS patient, that action
might be moving a cursor on a computer screen — a step toward reading and sending E-mail or surfing the Internet.
Williams and Garell are among the first to place the devices on the brain, rather than on the scalp (the skull "muddies" the brain's information). They are working with some of Garell's severe epilepsy patients, whose last resort is surgical removal of the tiny part of their brain in which their seizures originate. Garell temporarily implants sensors on the patients' brain to help pinpoint the seizures' origin. The patients remain hospitalized while doctors track their seizure activity.
Meanwhile, Williams and his students, who know in general the sensors' locations, work with the patients to learn more about which specific neurons are within a sensor's range. "We'll have them open and close their hand if we think it's over a hand motor area; move their arm, move their shoulder, move their leg, to try to map out the motor areas," says
Williams. "If it's in a sensory area, we'll play them sounds, give them images."
Once they locate a specific area of the brain that responds to a physical stimulus, the researchers do an "illusion screening." In other words, the patients think about moving their arm, or think about hearing a certain sound or seeing an image while Williams and his group record the neurons' response. "We'll pick the best area — the one that elicits the largest electrical activity in response to the illusion, as opposed to nothing," he says.
Within minutes, the epilepsy patients learn to use only their thoughts about certain motions, sounds or images to move a cursor on a computer screen. Just as quickly, they master a simple computer game — a version of the classic Pong.
These studies enable the research team to understand how quickly ALS and other patients might be able to learn the same skills. In addition, because each epilepsy patient's sensors are implanted in brain locations specific only to them, the researchers
can study which areas of the brain — in particular, non-motor areas — would be the best places to locate permanently implanted sensors for the motor-disability patients.
The researchers also are determining whether using their thoughts to communicate
via cursor might become so natural to the epilepsy patients that they could do it even if they were holding a conversation at the same time.
The work is an outgrowth of years of studies with monkeys. "We've shown that if we put some of these types of devices into monkeys, they can learn to do extraordinary things," says Williams. "They can learn to move cursors on a screen, move robotic arms around, and feed themselves with robotic arms. Our goal here is to translate some of these original ideas into the human patient population."