Mind Over Matter (Microchip Implant Trials to Begin)
Zina Moukheiber , 03.15.04

Once the realm of science fiction, neural prosthetics are slowly becoming reality.

In May a tiny, three-year-old company begins what could be a mind-bender of a medical-device trial. Five quadriplegics
will have a baby-aspirin-size microchip implanted under the skull, with a wire running to a nearby computer.
Neuroscientists will ask them to think about moving a cursor on a computer screen, and if the chips record their
thoughts correctly, the cursor will move by itself. The goal of the firm running the trial, Cyberkinetics of
Foxborough, Mass., is to help the severely paralyzed regain some function--dialing a phone, switching on lights--
through thought power.

Neural prosthetics have long been the stuff of science fiction. Morpheus and his crew from the Matrix series, and Case,
the hero of William Gibson's Neuromancer, all "jack" their brains directly into computers. But the science is catching
up with the fiction, and several companies are slowly making neural prosthetics real. "It's like the early days of the
cardiac pacemaker," says Timothy Surgenor, Cyberkinetics' chief executive.

Successful neural implants have thus far delivered mild corrective current to regions of the brain involved with motor
control. Medtronic's Activa device, which generates an estimated $100 million in annual sales, sends weak electrical
pulses from a stimulator near the collarbone to a pair of electrodes near the brain's thalamus region to calm the
severe tremors that afflict Parkinson's patients. Another company, Cyberonics, will probably generate $100 million in
revenue from a device that delivers jolts to the vagus nerve, reducing the frequency of epileptic seizures (FORBES,
Mar. 5, 2001). Northstar Neuroscience is expected to complete midstage clinical trials in March for a device that
electrically stimulates the brains of stroke victims to help them recover use of their limbs. Northstar, still
privately held, has raised $57 million in venture capital.

The next frontier is in going in the opposite direction, using implanted sensors to decipher neural intent and convert
it into movement of a cursor or, much further off, an artificial limb. Cyberkinetics' Surgenor is focusing on the
300,000 to 400,000 Americans he estimates are paralyzed from spinal cord injuries, Lou Gehrig's disease, cerebral palsy
or other maladies. Cyberkinetics has raised $9.3 million in venture money since its inception in 2001.

The human brain is a difficult computer to harness. We have 100 billion neurons, each connecting to 1,000 others.
That's 50 trillion roundtrip connections. Academics have been tinkering with neural sensors for decades, with notable
recent work done at Duke University, Caltech and the University of Pittsburgh. Philip Kennedy, the founder of Neural
Signals, rocked the science world with a successful human implant in 1998.

Cyberkinetics is one of the earliest to move the research beyond the campus. When its cofounder, John Donoghue,
chairman of Brown University's neuroscience department, began eavesdropping on the brains of rats and monkeys in the
early 1980s, the technology was still primitive and unreliable. Donoghue would glue a bunch of wires together and
insert them into an animal's frontal cortex, the region controlling movement. As the neurons fired off electrochemical
messages to one another, triggering responses in the muscles and other parts of the body, a computer would record the
amplified staccato chatter of, at most, two neurons. "We needed to study many brain cells at once," says Donoghue.

In 1994 Donoghue came across Richard Normann, who would later start a company called Bionic Technologies. Normann had
developed a 4mm silicon chip containing 100 microelectrodes, each able to pick up the signal of one or two neurons.

To prove they could translate thought into action, Donoghue and his team of three scientists started implanting
Bionic's chip in monkeys. The platinum tip of the electrode sat near the trunk of the neuron where bursts of signals
form. Donoghue connected the electrodes by wires to a computer and trained the monkeys to use a joystick to track a dot
on a computer screen with a cursor. He encouraged them to make as many hand movements as possible by rewarding them
with orange juice.

The Brown team then tried to match the hand movements to the monkeys' neural activity, measured by spikes on an EKG-
like printout. Three consecutive spikes could mean a move to the left, ten spikes a move to the right. Donoghue
recorded signals from up to 30 neurons, looking for patterns. Even though a hand movement is a result of millions of
neural signals, the output from as few as 7 to 30 neurons was enough to move a cursor. Donoghue then switched off the
joystick, and the code the humans had written allowed the monkeys to move the cursor with their brains.

Decoding human brains in this fashion seemed years off, until Phil Kennedy's 1998 groundbreaking procedure. Near the
surface of the brain of a paralyzed stroke victim he implanted a tiny cone-shaped device that wirelessly relayed neural
signals to a computer. The surgery took ten hours, and the patient was able to type by thinking, although it took him
20 seconds to get a letter out. It was primitive in that it only recorded one neuron.

Yet Neural Signals hasn't taken off. Kennedy has made only feeble attempts to raise venture money, funding the
technology on his own and with government grants totaling $1.7 million. "I'm not interested in selling out--this is my
baby, we'll grow together," says Kennedy.

Donoghue was more willing to tap outside investors. In 2001 he formed Cyberkinetics with a $75,000 grant from the state
of Rhode Island. Too busy to write a business plan, he enlisted Brown sophomore Mikhail Shapiro. Shapiro combed
academia for patents, negotiating exclusive licenses with Emory University, Brown and MIT.

In March 2002 Donoghue published the results of his monkey experiments in Nature. Mark Carthy, a venture capitalist at
Oxford Bioscience Partners, was captivated enough to consider investing. "The animal data were phenomenal," he says.

Before he would invest, Carthy wanted Cyberkinetics to own the factory that made the brain chips. After convincing
Bionic Technologies' founder to sell his company, which was already generating $500,000 in revenue, for an 8% stake in
Cyberkinetics, Oxford invested $4 million in Cyberkinetics for a 33% stake. The company has raised another $5.3 million
since. Shapiro, who got some equity, is now a senior at Brown doing his own research in neuroscience.

Cyberkinetics is awaiting Food & Drug Administration approval to begin its first human trials. Surgeons will take about
three hours to implant the 4mm chip into a hole on the surface of the cortex of the five quadriplegics. Wires will
emerge behind the ear to connect with a computer.

Some are skeptical. "The technology is a little premature to commercialize," says Duke neuroscientist Miguel Nicolelis,
who showed that monkeys can move robotic arms remotely by thinking. "We have to be very careful. It's like gene
therapy. If we make one mistake, the whole field is dead."

SOURCE: Forbes
http://www.forbes.com/home_asia/free_forbes/2004/0315/186.html

* * *

----------------------------------------------------------------------
To sign-off Parkinsn send a message to: mailto:[log in to unmask]
In the body of the message put: signoff parkinsn