As with about a quarter of locked-in patients, Hawking has retained a tiny bit of movement, the shifting of his eyes. Yet he manages to communicate by focusing on different points of a computer screen filled with different letter combinations. He can form words, and so has been able to continue his work in explaining black holes, the origin of the universe and its future.
Most locked-in patients, however, cannot communicate with their eyes as well as Hawking.
The electrode implanted under M.H.'s skull, atop the cortex, only picked up increased or decreased brain activity (Kennedy's later implants would pick up more detailed signals). It amplified the electric signals by about 1,000. The signals were transmitted to an FM radio receiver attached to the skull and then sent as radio waves to a computer across the room.
To test the electrode, Kennedy streamed M.H.'s brain activity through the computer, playing the noise through a speaker. He likens the sound to "running a stick down a railing."
"We said, 'Listen to the signals. We want you to quiet it down. Just stop everything.' And then we said, 'Just make it really active.' And she did," Kennedy recalls. "That's all it was. It was very crude. But it was very effective. It showed she could affect the signals."
She died 76 days after the experiment, from pneumonia and kidney failure. Softening and strengthening her brain signals were her last communications with the world.
Then came Johnny Ray, a drywall contractor who played jazz guitar and who'd been locked into his body at age 43 by a stroke.
At the time Ray received his implant in March 1998, he could still move his face a little, smile and shift his eyes. Kennedy implanted two electrodes next to the part of Ray's brain that controlled his left hand. He started asking Ray to try communicating with the computer by thinking about moving the mouse with his hand. The cursor moved.
Nine months into Ray's training, which was so exhausting it could not exceed 20 minutes a day, Kennedy asked him to spell his name by moving the cursor over a screen of letters. Ray spelled JOHN on two of his first four tries. He took a break, tried again and didn't do so well, presumably because of fatigue. He spelled JOHLQQQ.GYUVWABDN, then HIJJROHNLN, then JOIH.N. "And then he started to spell P," Kennedy says. "And I thought, 'Oh well, I'll just leave him alone.
"And then he spelled my name." PHIL.
Ray's mastery of the computer improved over the next six months. On May 24, 1999, Kennedy visited Ray at his home in the V.A. Medical Center in Decatur. Kennedy asked what he felt as he moved the cursor. Ray answered NOTHING. He had learned to move the cursor simply by thinking of moving the cursor. He no longer had to think about moving his hand. The part of his brain that was supposed to dictate movement evolved; it abandoned hope for the appendage. The brain opted instead to communicate directly with the computer.
"That led us to say he has now devoted that part of his brain to driving the cursor," Kennedy says. "See, the brain is very crafty. It's very adaptable. It can learn. And that's really the key thing."
Three months later, Kennedy moved the homeless Neural Signal Inc. into Advanced Technology Development Center, which is affiliated with Georgia Tech. One of the air-conditioning vents drips when it rains ("What do you expect for cheap rent?" Kennedy says). The sign on the door says, "More Power From Your Brain."
Three years after that, Ray died from a brain aneurysm.
"He was the world's first cyborg," Kennedy says. Ray was also Kennedy's star student, leaving an indelible mark on neuroscience. He proved that man can train the neurons in his brain to do more than what they were created to do. Man can command his neurons, and that made the simplicity of Kennedy's design all the greater. He now believes that to re-wire a paralysis victim, a handful of electrodes strategically placed near a few good neurons could do the job.
"A lot of people now agree that you don't need hundreds and hundreds of neurons, but just a few good ones," Kennedy says. "If we can hold neurons that long and then control their activity, [future patients] will learn to use those signals to control what they want."
Even muscle activity, Kennedy supposes. His next big goal is to apply the same technology to the parts of the brain that control motion. "What we're trying to do is to [implant] ALS patients early, before they lose the ability to speak," Kennedy says. That way they can train. "We had a few lined up, but we're still looking for patients."