By wiring into brain neurons, scientists have enabled rats to control a mechanical arm without lifting a paw. The feat, reported in the July Nature Neuroscience, may pave the way for neuroprosthetic devices for paralyzed patients.
Spinal cord injuries and neurodegenerative diseases can disrupt the brain's control of muscles, leading, in severe cases, to total paralysis. Researchers in the past have been able to hook up individual brain neurons in animals to electrodes, but the resulting signals failed to convey enough information to operate a prosthesis. Neuroscientist John Chapin and his colleagues at the Medical College of Pennsylvania Hahnemann School of Medicine in Philadelphia were keen to try routing signals from enough brain neurons to make a simple device work.
Toward that end, Chapin's team devised an array of up to 48 electrodes that could be implanted in rats. First the researchers trained the animals to press a lever controlling a robotic arm. By adjusting the pressure on the lever, the animals, when thirsty, could direct the arm to collect water droplets from a dripping faucet. How far the arm swung, the researchers found, corresponded to the intensity of the overall neuronal activity in the motor cortex and in the thalamus, regions that control voluntary limb movement. Nerve firing peaked shortly before the rats pressed the lever.
The researchers next amplified signals from these brain regions into electrical pulses that could drive the robotic arm. The big test came when they disconnected the lever from the arm. "After a few days the rats figured out that they didn't have to press the lever" to get the water, Chapin says. In about 25% of the attempts, the animals could move the robotic arm simply by firing the neurons that normally control their paw movement, Chapin says.
"It's an interesting demonstration" that could pay off for patients someday, says Eberhard Fetz, a neuroscientist at the University of Washington in Seattle. He adds, however, that "a robotic arm with just a single hinge is a fairly simple control problem." An arm useful to paralyzed people would require a far more sophisticated device, Fetz says.