Brown University scientists are featured in the journal Nature for developing a brain implant that enables paralyzed people to control an object simply by thinking about it.

01:00 AM EDT on Thursday, July 13, 2006

BY FELICE J. FREYER
Journal Medical Writer

In the neurology clinic at Massachusetts General Hospital, Dr. Leigh R. Hochberg sees people with severed spinal cords who cannot move their arms and legs, people with muscles wasted by disease, and stroke victims who are "locked in" -- conscious but unable to speak or move.

Today Hochberg can offer them little more than support and comfort. But he has a hope: "I hope that one day we'll be able to say, 'We have a technology that will enable you to move again.' "

Hochberg, who is also an investigator in neuroscience at Brown Medical School, is not just dreaming. He's working on that very technology: a brain implant and computer system that has already enabled four paralyzed people to control an object simply by thinking about it.

New and cumbersome, the system still has a long way to go. But experimental subjects have opened e-mail, played computer games, adjusted the volume on a TV set, and manipulated a robotic arm. If the device were connected to a system that stimulates muscles, someday such people might be able to use it to move their limbs.

Hochberg is the lead author of an article in today's issue of the journal Nature, describing how a team of researchers at Brown University put the technology to work with the first of their experimental subjects.

Called BrainGate and made by Cyberkinetics Neurotechnology Systems in Foxboro, Mass., the device has previously been described at scientific meetings and in the media. But the Nature article marks a milestone, detailing the findings for the medical and scientific community -- and featuring Brown researchers on the cover of a prestigious peer-reviewed journal.

"It gives it a level of credibility that is very important for this kind of project," said Dr. Jon A. Mukand of the Sargent Rehabilitation Center in Warwick, who is one of the article's authors. The fact that more than one patient has successfully used the system means that "people in the disabled community can feel more optimistic," Mukand said.

In June 2004, Rhode Island Hospital surgeons implanted a tiny computer chip onto the brain of Matthew Nagle, a 25-year-old Massachusetts man whose spinal cord had been severed when he was stabbed in the neck. Nagle cannot move his arms or legs and relied on a ventilator to breathe. The chip, it was hoped, would pick up signals from Nagle's brain when he thought about moving his arm, run those signals through a "decoder," and translate them into commands a computer could understand.

A similar system had worked well in able-bodied monkeys. But the researchers didn't know whether the brain of a man who had not moved his arms and legs for three years could still produce movement signals. They didn't know whether the chip they'd designed -- with 96 hair-thin electrodes -- would pick up enough information, nor whether those signals could be successfully routed to a personal computer. They didn't know whether the chip would leave the patient's brain unharmed.

Today's article offers affirmative answers to all those questions.

"It's a huge step forward," said John P. Donoghue, the neuroscientist whose Brown laboratory developed the technology during more than a decade of research in monkeys. "We didn't know whether in people it would work -- especially people who had destruction of their nervous systems."

When the implant in Nagle's brain was hooked up to the computer system, Nagle would think about using his arm to move a computer mouse. The part of his brain that, before his injury, had controlled his arm was indeed still firing off signals, and the cursor moved. He drew squiggly circles, opened fake e-mail and played Pong. He learned to operate a prosthetic arm by thinking about arm movement. He got so good that he could grab an object with the arm and drop it in the hand of a researcher.

Nagle kept the implant for nine months with no ill effects. His brain waves were tested shortly before its removal and were found to be normal. "The brain accepted it very well," Donoghue said. "It's removable and replaceable."

Since then, two other people with spinal cord injuries, and one person suffering from Lou Gehrig's disease, or amyotropic lateral sclerosis, have entered the study. According to Donoghue, who is director of Brown's Brain Science Program, they have had results similar to Nagle's.

But many obstacles remain. The Nature article noted that signals from the implant declined after six and a half months, and the implant has never been used for more than 14 months. Researchers need to develop an implant that can function continuously and safely for many years.

The system is also too unwieldy to be practical. Nagle operated it through a soup-can-like device screwed to the top of his head and wired to a cart full of electronics. For each session, an attendant spent an hour hooking up the apparatus before Nagle could use it.

The ultimate goal is to develop a wireless, miniature version that can be fully implanted in the patient's brain, Donoghue said. This is an engineering challenge that Donoghue believes is achievable within years, thanks to rapid advancements in the miniaturization of electronics. "There's some technical challenges," Donoghue said, "but it's not a big deal. It can be done."

Another goal is to improve the patient's control of the cursor by fine-tuning the software that translates the brain signals. In moving a cursor to a target on the screen, Nagle's accuracy was 75 percent to 85 percent.

In a separate article in today's Nature, researchers at Stanford University report that they've developed a new method of decoding the brain's signals that works four times faster than existing technology. Donoghue said their work is potentially useful and could be adapted to his system. He said that the Stanford method was like using a keyboard, while his was like using a mouse. Anyone operating the system he envisions would probably want both, he said.

The ability to operate a computer would give a paralyzed person tremendous control over the environment, whether by commanding a household robot or merely turning on the lights. But it might also go beyond that someday, if the brain implant can be connected to systems that stimulate muscles to move. Such systems are under development at Case Western Reserve University in Cleveland, and Donoghue's team has a new federal grant to work with researchers there.

"You basically patch together the nervous system," he said.

"The real dream for this technology," said Hochberg, "is to reconnect brain to limb."

The study of BrainGate in humans was financed by Cyberkinetics, a private company that Brown researchers founded in 2001 to bring their work to market. Researchers around the country have been testing similar devices in monkeys, and one group has implanted a simpler electrode in three humans. But the Cyberkinetics trial is the only research project testing a 96-electrode implant in people.

Another approach prevents the hazards of surgery by using electrodes on the scalp that tap into brain waves, but these systems require considerable learning and concentration. (Nagle, in contrast, was able to carry on a conversation while moving the cursor or manipulating the prosthesis.)

"John Donoghue's lab, which has done a tremendous amount of fundamental science with monkeys, is really one of the pioneering labs to show similar things in humans. Their translation to humans has been really fantastic," said Krishna Shenoy, a Stanford engineer and neuroscientist, and an author of the Stanford paper. "Reports like the Donoghue report really offer a lot of new hope to severely paralyzed patients. This is not to build up expectations too much. It's not going to be tomorrow."

ffreyer@projo.com / (401) 277-7397