The findings also have implications for the development of neural prosthetics. For example, the device could be connected directly to the spinal cord or muscle rather than to another part of the brain. "If a person had a spinal-cord injury and the link from brain to muscle is impaired, this connection could bypass that injury and reconnect brain cells to the muscle," says Fetz. His group is currently working on this application.

Such a device might one day be used to boost the effects of rehabilitation therapy. Rehab exercises are designed to boost the brain's innate plasticity--in essence, they try to make the damaged brain develop a new neural pathway to control movement of specific muscles. Devices such as Fetz's chip may be able to speed along this process by strengthening connections between two different brain areas. "This study gives a preliminary indication that there are methods that can be used to almost engineer this rerouting," says Andrew Schwartz, a neuroscientist at the University of Pittsburgh who studies neural prosthetics.

While the findings are exciting, there is still a long way to go before the technology can be applied to human beings. In terms of stroke rehabilitation, scientists would first need to figure out precisely which two parts of the brain should be reconnected. "In order to restore function, you can't just make more connections--you have to make the right connections," says John Donoghue, a neuroscientist at Brown University who is developing a different type of implantable prosthetic.

Donoghue's chip, which is already being tested in human trials, uses many recording electrodes, but it currently doesn't have the ability to stimulate other parts of the brain or body. (With his device, neural signals are sent to a computer, which decodes the information and uses it to move a cursor on a computer screen. See "Implanting Hope," March 2005, and "Brain Chips Give Paralyzed Patients New Powers.") However, Donoghue says he is currently working on stimulating capabilities as well.