CHICAGO—Innovative neurotechnology may give paralyzed patients with spinal cord injury, stroke, and motor neuron disorders greater independence and control over activities of daily living. With the use of a brain-computer interface system, John Donoghue, PhD, and colleagues showed that four tetraplegic patients were able to voluntarily modulate neural activity patterns in the motor cortex through imagined actions, despite having different forms of CNS impairment. One participant, unable to move or speak because of a brain stem stroke, learned to control a communication device and remotely operate a wheelchair using her thoughts.

"This is a particularly exciting time in the history of neurology and medicine," said Dr. Donoghue, at the 131st Annual Meeting of the American Neurological Association. "There are a number of neurotechnologies now emerging that will provide a whole range of new tools that will be available to treat and diagnose neurologic disorders." Dr. Donoghue is a Professor of Neuroscience and Director of the Brain Science Program at Brown University in Providence, Rhode Island, as well as a founder and Chief Scientific Officer of Cyberkinetics Neurotechnology Systems, Inc.

BRAINGATE

"The concept behind this technology is simply to take the signals directly from the motor cortex—for example, from the arm area of the motor cortex—and bring them to the outside," said Dr. Donoghue. "The current system has a percutaneous connector that connects to a computer cable to bring the signals outside, to a cart that contains computers that amplify, process, and interpret the signals. The brain’s ‘intention to move’ is represented—in the simplest case—as a cursor on the computer screen. In other words, what you would ordinarily do with your hand—move the computer mouse—you would be doing by just thinking about moving it.... We actually detect and decode the intended motion of the hand by analyzing patterns of motor cortical activity that are recorded by the sensor. Numerous labs have shown in monkeys that you can, in fact, decode the brain activity well enough that this signal can substitute for the hand to control a computer cursor."

One of the first issues to be addressed in using the technology was whether neural signals related to movement were still present in paralyzed people. "Much to our delight, as soon as we turned on the device, we saw neural activity," noted Dr. Donoghue. "Even years after an injury, in all of our patients, we see that the neural spiking activity remains.... Using electrode arrays, we can record many signals in real time in the motor cortex."

MIND OVER MATTER

Recent work has shown that cursor control can be substantially improved."Not only do we get a smoother motion; [participants] can also select ‘click,’ using only neural activity in the motor cortex," commented Dr. Donoghue. "So we can sort of emulate your computer mouse function, and I think we can do a pretty good job of that, although it is slightly slower than you or I would do it."

In another demonstration, one participant used a cursor to sweep over an icon to open e-mail or draw. "Although that quality of control in this early demonstration is nothing like what we have, it was similar to what maybe a 2- or 3-year-old child would have," said Dr. Donoghue. "We’re basically asking and using that crude filter to get control using a few dozen neurons when normally we use millions of neurons. Advances in the mathematical decoding strategies have led to significant improvements in the quality of control."

The investigators also tried to show whether individuals could learn to control physical devices and control them without using a computer interface. With the aid of a motorized prosthetic hand, Dr. Donoghue’s group decoded a patient’s brain signals and used them to control the motor that opened and closed the hand. "We just had him look at the hand," explained Dr. Donoghue. "Now he’ll say ‘open’ and ‘close,’ and you’ll see the action of the hand as he imagines that. What is significant about this is that it could allow him to manipulate objects so effectively that he can interact with this world for controlling devices."

The patient was also able to use a simple toy robot as an interface to close a type of hand gripper, grabbing a piece of candy and dropping it in the technician’s hand. "It took him about seven minutes from the time we set up the robot for him to get it," said Dr. Donoghue. "So he quickly integrates how to use this neural activity to do something useful. Again, these are very simple demonstrations. They need a lot of work before they are something that is a practical device."

"Ultimately, we hope that there would be a wide range of ways that we can connect the motor cortex up to wheelchairs, computers, and other devices," said Dr. Donoghue. "But I think one of the things that is really exciting is the ability to hook this back up to the muscles again in spinal cord–injured patients." He and his colleagues are also developing a neural prosthesis that uses the BrainGate system to actually control the arm and hand of a person with tetraplegia. "We’re not saying that this person is going to move in what is considered a normal way," Dr. Donoghue pointed out. "Our goal is to get that person’s arm supported in an orthotic device, to have the hand open and close under neural control and grab, say, a cup and bring that cup to the mouth. If we do that within five years, we will be very excited.

"We are on a path to actually reconstruct the nervous system in [a process] that sends the brain signals back to the muscles again," Dr. Donoghue continued. "But I think that really is an ambitious and very long-term goal. But we see some day that there would be this vision of a bidirectional interface of physical components with the nervous system that are able to restore sensory input and motor output. In addition, I think we can allow patients to participate in their own rehabilitation, that is, motion of the limb, through an exoskeleton that is neurally controlled…. I could further envision one day implanting neural interface technology, placed at or near the location of an epileptic seizure, to report that a seizure is going to happen. Neurons are very sensitive measures of what is going on in the brain. By predicting that a seizure was about to happen, a patient could, for example, get off the road, or take medication to reduce or prevent the seizure. Or, connect it … in a way to deliver medication directly to the epileptic site. I think the promise of the future is extremely bright, and I think this is just the very beginning of this whole field."

NR