GIVING a high-five. Rubbing his girlfriend's hand. Such ordinary acts - but a milestone for a paralyzed man.

True, a robotic arm parked next to his wheelchair did the touching, painstakingly, palm to palm. But Tim Hemmes made that arm move just by thinking about it.

Emotions surged. For the first time in the seven years since a motorcycle accident left him a quadriplegic, Hemmes was reaching out to someone - even if it was only temporary, part of a month-long science experiment at the University of Pittsburgh in the state of Pennsylvania.

"It wasn't my arm but it was my brain, my thoughts. I was moving something," Hemmes says. "I can't describe it."

The Pennsylvania man is one pioneer in a quest for thought-controlled prosthetics to give the paralyzed more independence - the ability to feed themselves, turn a doorknob, hug a loved one.

The goal is a Star Trek-like melding of mind and machine, combining what's considered the most humanlike bionic arm to date - even the fingers bend like real ones - with tiny chips implanted in the brain. Those electrodes tap into electrical signals from brain cells that command movement. Bypassing a broken spinal cord, they relay those signals to the robotic third arm.

This research is years away from commercial use, but many teams are investigating different methods.

At Pittsburgh, monkeys learned to feed themselves marshmallows by thinking a robot arm into motion. At Duke University in North Carolina, monkeys used their thoughts to move virtual arms on a computer and got feedback that let them distinguish the texture of what they "touched."

Through a project known as BrainGate and other research, a few paralyzed people outfitted with brain electrodes have used their minds to work computers, even make simple movements with prosthetic arms.

But can neuroprosthetics ever offer the complex, rapid movements that people need for practical, everyday use?

"We are at a tipping point now with this technology," says Michael McLoughlin of the Johns Hopkins University Applied Physics Laboratory, which developed the humanlike arm in a US$100 million project for DARPA, the Pentagon's research agency.

Pittsburgh is helping lead a closely watched series of government-funded studies over the next two years to try to find out. A handful of quadriplegic volunteers will train their brains to operate the DARPA arm in sophisticated ways, even using sensors implanted in its fingertips to try to feel what they touch, while scientists explore which electrodes work best.

"Imagine all the joints and motions in your hand," says Pittsburgh neurobiologist Andrew Schwartz. "It's not just reaching out and crudely grasping something. We want them to be able to use the fingers we've worked so hard on."

The 30-year-old Hemmes' task was simpler. He was testing whether a new chip, which for safety reasons the Food and Drug Administration let stay on his brain for just a month, could allow for three-dimensional arm movement.

He surprised researchers the day before electrodes were removed. The robotic arm whirred as Hemmes' mind pushed it forward to hesitantly tap palms with a scientist. Then his girlfriend beckoned. Hemmes slowly raised the black metal hand again and slowly rubbed its palm against hers a few times.

These emotional robotic touches have inspired researchers now recruiting volunteers for soon-to-start yearlong experiments.

"It was awesome," says normally reserved Dr Michael Boninger, rehabilitation chief at the University of Pittsburgh Medical Center. "To interact with a human that way ... This is the beginning."

Hemmes' journey began in 2004. He owned an auto-detailing shop and rode his motorcycle in his spare time. One night he swerved to miss a deer. His bike struck a guardrail. His neck snapped.

His determination didn't. Paralyzed below the shoulders, he's tried other experimental procedures in hopes, so far unrealized, of regaining some arm function.

"Your legs are great ... but they just get you from here to there," Hemmes says near his home north of Pittsburgh. "Your arms and fingers and hands do everything else. I have to get those back."

His ultimate goal is to hug his 8-year-old daughter. "I'm going to do whatever it takes, as long as it takes, to do that again."

Hemmes entered an operating room at UPMC with a mix of nerves and excitement. "It's good anxiety," he says. "There is so much riding on this."

It seems so simple.

Think "I want that apple," and your arm reaches out and grasps it. You're not aware that neurons are instantaneously firing in patterns that send commands down the spinal cord - make the shoulder raise the arm, extend the elbow, flex the wrist and all five fingers.

A similar firing occurs when you imagine movement or watch the movement you'd like to perform, says Boninger, who with Schwartz is leading the Pittsburgh research bio-engineers, neuroscientists and physicians.

The DARPA arm was developed for amputees. Paralysis poses a bigger challenge: getting signals around a broken spinal cord.

Miniature electrodes tested

For quadriplegics, scientists use implanted electrodes, called a "brain-computer interface," to record electrical activity. Signals move down through wires that tunnel under the skin and out by the collarbone, and are plugged into a computer or a robotic arm.

Until now, researchers mostly have tested miniature electrodes that poke inside the brain's motor cortex and record from individual cells, presumably allowing for precise movements. Pittsburgh's next test-patient will have two penetrating grids implanted in different parts of the cortex for a year to record from 200 cells altogether.

In contrast, Hemmes' chip sat on the surface of his motor cortex, a less invasive method that records from groups of cells. The size of two postage stamps, it's based on Brelectrical signal mapping used to track seizures in epilepsy patients.

Both approaches need study, says Chen Daofen of the National Institutes of Health, who oversees neurorehabilitation research. He compares the options to eavesdropping on a party by sending in individual microphones or setting up a recorder at the window.

Hemmes' operation took two hours. He had practiced imagining arm movements inside brain scanners, to see where the electrical signals concentrated. That's where neurosurgeon Elizabeth Tyler-Kabara cut, attaching the chip through an inch-wide opening on the left side of Hemmes' skull.

Two days later, Hemmes was hooked to a computer, making simple cursor movements. The next week, he tried to trigger real movement using the arm.

Hemmes reclined in his wheelchair, the robot arm bolted to a steel rod nearby. The task: make the arm reach out and grasp a ball on a board. The arm whirs forward, stops, goes again, then pulls back.

"It's doing the opposite of what I ask it," Hemmes says in frustration. "When I think about reaching back, it goes forward."

Then he focuses on his elbow.

Hemmes takes a deep breath and tries. The arm whirs forward, reaching the ball. The fingers clench around it.

"There's no owner's manual," Hemmes says. "I'm training my brain to figure how to do this."

Letting go is harder, the motor growling as the arm tugs backward before the fingers release. Hemmes starts imagining his hand relaxing before pulling back, and the robot hand follows.

Sure, a robotic hand is useful but there's still no sense of touch.

Recreating sensation means crafting a two-way highway with brain chips. That's what Duke University, in a study published last week in the journal Nature, did with its two monkeys. When the animals "touched" objects on a computer screen with their video game-like arms, electrical signals flashed to implanted electrodes - different signals for different textures, to tell objects apart.

There's a plan for one Pittsburgh study patient to begin testing touch capability next year, with a similar attempt at the California Institute of Technology to follow.

What about moving paralyzed limbs? Duke's plan is to turn its research into a robotic exoskeleton to help the paralyzed move.

Hemmes is more intrigued by what's called functional electrical stimulation, zapping muscles with electrical currents to make them move. Researchers are working on that approach.