Brain Implants Help Paralyzed Monkeys Get a Grip

Spinal cord injuries cause paralysis because they sever crucial communication links between the brain and the muscles that move limbs. A new study with monkeys demonstrates a way to re-establish those connections. By implanting electrodes in a movement control center in the brain and wiring them up to electrodes attached to muscles in the arm, researchers restored movement to monkeys with a temporarily paralyzed hand. The work is the latest promising development in the burgeoning field of neuroprosthetics.

In recent years, scientists have taken many steps toward creating prosthetics to help paralyzed people interact more with the world around them. They've developed methods to decode signals from electrodes implanted in the brain so that a paralyzed person can control a cursor on a computer screen or manipulate a robotic arm with their thoughts alone. Such brain implants are still experimental, and only a handful of people have received them. Several hundred patients have received a different kind of neural prosthetic that uses residual shoulder movement or nerve activity to stimulate arm muscles, allowing them to grasp objects with their hands.

The new study combines these two approaches. Neuroscientist Lee Miller of the Northwestern University Feinberg School of Medicine in Chicago, Illinois, and colleagues implanted electrode grids into the primary motor cortex of two monkeys. This brain region issues commands that move muscles throughout the body, and the researchers positioned the electrodes in the part of the primary motor cortex that controls the hand, enabling them to record the electrical activity of about 100 neurons there. In a separate surgery, the team implanted up to five electrodes in three arm muscles used in grasping objects with the hand. By recording simultaneously from the brain and muscle electrodes as the monkeys gripped various objects, Miller and colleagues developed computerized decoding algorithms that predicted how signals from the brain translated into electrical activity in each of the three muscles. Next, they hoped to demonstrate that these algorithms could interpret commands from the brain in a paralyzed monkey and provide just the right stimulation to the muscles to elicit the intended movement.

To test this idea, the researchers injected a nerve-blocking drug to temporarily paralyze one hand and forearm in the monkeys. The animals were rarely able to do a simple task they'd previously learned, picking up a rubber ball and dropping it into a tube to earn a reward of juice. But with the neuroprosthesis switched on, the monkeys succeeded about 80% of the time, the team reports online today in Nature.

"What's fundamentally different about our system is that we're going back to the brain and eavesdropping on the signals that normally occur," Miller says. He hopes this type of neuroprosthesis will eventually enable paralyzed humans to move a hand just by thinking about it. That would be an improvement over current muscle-stimulating prostheses, which rely on residual movement in other body parts and require patients to learn unnatural movements such as moving a shoulder up and down to grasp and release. "There really are not major technical hurdles to trying this in humans on an experimental basis," Miller says, but he adds that getting approval and recruiting patients for such a trial will probably take a few years.

The work is a first step toward integrating brain decoding and muscle stimulation, says Andrew Schwartz, a neuroscientist at the University of Pittsburgh in Pennsylvania. But he cautions that spinal injuries often cause more widespread paralysis than Miller's monkeys had. "In a truly paralyzed arm, many more muscles would have to be activated," Schwartz writes in an e-mail. "To achieve more natural behaviors, complex combinations of muscles would need to be activated, and this kind of control is much more complex than what was shown in this demonstration."

Miller agrees that achieving greater dexterity is a major challenge. One approach his team is investigating is to stimulate the nerves that control muscles in the arm and hand instead of the muscles themselves. "If we can work out the details," he says, "in principle we should be able to activate all the muscles that those nerves innervate."