Five years ago, a young man who could not move his legs or arms learned to control a computer cursor by imagining his hand manipulating a mouse.

Matthew Nagle, paralyzed after being stabbed in the neck, could open e-mail, change the TV channels and draw a very wobbly circle on the screen. He also moved a prosthetic arm by thinking about moving his own arm.

Nagle could accomplish those tasks because doctors had implanted a silicon wafer with 100 hair-thin electrodes in his brain, and this sensor was connected to an elaborate computer system that interpreted his brain signals. Called the BrainGate Neural Interface System, this invention was developed at Brown University.

Nagle, who later died of causes unrelated to the experiment, was one of four paralyzed people who tested BrainGate. The others were another person with a spinal cord injury, a man with Lou Gehrig’s disease and a woman totally immobilized by a brain-stem stroke.

But last year the private for-profit company that financed the BrainGate study, Cyberkinetics Neurotechnology Systems, backed away from the project and eventually shut down, a victim of the faltering economy and the extremely high cost of conducting such research.

Now, the researchers are announcing Wednesday morning that they have cobbled together enough federal and philanthropic money to sustain their research and are recruiting 15 patients for a second, larger clinical trial. They estimate they have won more than $8 million in grants from the Department of Veterans Affairs, several branches of the National Institutes of Health and three charitable foundations.

John P. Donoghue, director of the Brown Institute for Brain Science, and Dr. Leigh R. Hochberg, a Brown engineering professor and a neurologist at Massachusetts General Hospital, are directing the research, which will be based at Mass. General. People who cannot use their hands –– most will be totally paralyzed –– and who live within two hours of Boston or Providence can apply to participate in the BrainGate2 pilot study.

As with the first study, the purpose of BrainGate2 is to make sure the device is safe; researchers are also learning how well it functions and making improvements. They hope eventually to enable paralyzed people to control their environment through assistive devices and –– possibly someday –– control their own muscles with electrical stimulation.

The results from the first study were promising, Donoghue said. None of the participants was harmed by the device despite many months of use –– a total of more than 2,800 days for all four. All were able to control the computer cursor, although their accuracy varied from day to day. Most significant, the BrainGate research showed that even people who had not moved their bodies for years still had activity in the part of the brain that controls motion. So the device was able to pick up signals when they thought of motion.

Although the participants kept the implant in their heads, they could use it only once or twice a week when the researchers came to connect them to a cartload of computers.

Nagle died of a blood infection in 2007, long after the device was removed. In the other spinal-cord-injured patient, a part of the BrainGate system malfunctioned and his participation was cut short. The patient with amyotrophic lateral sclerosis (Lou Gehrig’s disease) had the device for nine months, and died in an unrelated incident when his ventilator was accidentally disconnected.

The third patient, the woman with the brain-stem stroke, still has the BrainGate electrode array in her head after 42 months. Recently, Donoghue says, she eagerly signed up for BrainGate2.

In the second trial, the implant remains the same but the electronics are more advanced. Participants will be able to both point and click. Researchers are working on teaching them to “type” on a virtual keyboard, and on understanding why their control of the cursor is better on some days than others.

Donoghue helped found Cyberkinetics, and says he’s sad that it failed. “Cyberkinetics accelerated this field tremendously,” Donoghue said. “I figure it would have taken me 20 years to get the kind of money they provided.”

Other research teams around the country are also working on devices to allow the brain to control computers. Those being tried in humans sit on the surface of the brain, while other types of implanted devices are being tested in monkeys. Donoghue says his team is the only group testing such devices implanted in the brains of humans.

To inquire about the research, send an e-mail to clinicaltrials@braingate2.org. For more information, go to http://www.braingate2.org.

ffreyer@projo.com