Subscribe to eTOC

Hand Function Restored in Paralyzed Monkeys Through Brain-Machine Interface


Investigators were able to restore voluntary control of hand movements in monkeys that had been temporarily paralyzed using a noninvasive electrical stimulation technology.

Researchers have for the first time restored limited functional control of the hand by eavesdropping on the same motor neurons that normally control hand movement in monkeys, and then shuttling that information along to electrodes that control the necessary muscles through functional electrical stimulation (FES).

Previous efforts have used an implanted array of electrodes in the brain to control a robotic limb or computer cursor rather than the paralyzed hand itself. Another approach, already on the market, requires users to move their shoulders in a particular way to activate pre-programmed stimulation of paralyzed muscles in the hands, enabling them to grasp or release.

The new effort, described in a Nature paper published online on April 18, is simpler and less cognitively demanding for users, as the end result — hand movement —is achieved through normal cognitive intention.

“Granted it's still in a monkey, but it's a very exciting proof of principle that subjects are able to integrate this brain-machine interface with natural movements of the intact muscles in the same arm...and that they're able to use this to reanimate the paralyzed muscles,” said Marc Slutzky, MD, PhD, assistant professor of neurology, physiology and physical medicine at Northwestern Feinberg School of Medicine. Dr. Slutzky was not involved in the current study, but had collaborated previously with the investigators.


For the study, two rhesus monkeys were trained to pick up weighted rubber balls and to convey them to an opening at the top of a dispenser. An array of 96 microelectrodes was then implanted into the hand area of their primary motor cortex (M1), while intramuscular electrodes were implanted in a separate operation for recording and stimulation of hand and forearm muscles.

After observing the activity of the neurons and the corresponding activity of the hand muscles during the monkeys' performance of the trained task, the researchers then simulated some of the effects of paralysis caused by C5 or C6 spinal cord injury by injecting the monkeys with a local anesthetic to block median and ulnar nerves at the elbow. They then recorded signals from the brain and used the information gathered while observing normal brain and hand muscle activity to predict the intended activity of several of the paralyzed muscles, including the degree of intensity of muscle activity. The monkeys were then able to successfully manipulate the ball as if their hands were not paralyzed.

“This process essentially bypassed the spinal cord, restoring to the monkeys voluntary control of their paralyzed muscles,” wrote the team led by Lee E. Miller, PhD, professor of neuroscience at the Northwestern University Feinberg School of Medicine. “This achievement is a major advance toward similar restoration of hand function in human patients through brain-controlled FES.”


“The remarkable thing is they're putting together two complex systems, the neural control and the FES, and showing that there's a potential for both to restore function,” said Karunesh Ganguly, MD, PhD, assistant professor of neurology at the University of California, San Francisco. “Putting them together is a great proof of principle that these two complex systems can work together,” added Dr. Ganguly, who is studying similar brain-machine interfaces as a treatment for paralysis.

Even so, Dr. Ganguly said, translating the system directly into paralyzed humans as a long-term solution will be challenging.

“Part of the problem is that you inevitably have these electrodes moving around a bit inside the brain, moving away from the neurons they're recording from, and probably bleeding a bit in the process,” he said. “There is also an inflammatory reaction. The end result is you eventually see scarring and the electrode loses its ability to conduct electricity.”

Rather than work with implanted microelectrodes that record the activity of single neurons, Dr. Ganguly studies the use of larger electrodes placed above the dura to record the activity of a field of neurons simultaneously. While such information is less discrete, he said, “It's a technology that's more stable. From a translational point of view, it's probably going to be the first step.”

DR. LEE E. MILLER: “A deep-brain stimulator for a Parkinsons patient goes down multiple centimeters into the brain. The electrode array we implant penetrates just 1.5 millimeters into the cortical surface. Its much less invasive. In my mind, theres a greater potential for this procedure than for deep brain stimulation.”

Dr. Slutzky agreed that recording from multiple neurons at once from outside the brain will likely prove more feasible than recording from individual neurons with implanted microelectrodes.

“You lose some signal quality, but the question is: do you need that much signal quality?” he said. “Several groups, including my own, are looking at potentials from multiple neurons, sometimes called field potential or multi-unit potentials, to see if we can do just as well with those. One of our findings is that it seems we can do almost as well as with single neurons.”

In an interview with Neurology Today, Dr. Miller said the implantation of the microelectrodes is safer than those used in deep brain stimulation.

“A deep-brain stimulator for a Parkinson's patient goes down multiple centimeters into the brain,” he said. “The electrode array we implant penetrates just 1.5 millimeters into the cortical surface. It's much less invasive. In my mind, there's a greater potential for this procedure than for deep brain stimulation.”

Although he has not yet begun human trials, Dr. Miller said he hopes it will be possible, and that Leigh R. Hochberg, MD, PhD, has already implanted a similar array into the brains of half a dozen or so paralyzed humans.

In an e-mail, Dr. Hochberg, who has appointments at both the Brown University School of Engineering and Harvard Medical School, said: “The Miller Lab's nice recent results, obtained in a meaningful preclinical model, provide further proof of concept that cortically-controlled FES has the potential to restore intuitive and useful hand grasp.”

Another barrier to bringing the implanted microelectrode array out of the research laboratory and into general use is that existing models have wires from the microelectrodes emerging out of the skull.

“It has to be plugged in,” said Dr. Slutzky. “Until it becomes wireless, that's the biggest limiting factor. Even the trials have some difficulty recruiting patients. One of the patients said she didn't want her kids seeing her with a plug sticking out of her head. But a lot of people are working on making it wireless. So I certainly think that within the decade there's a possibility for a clinical trial of this kind of system.”


• Ethier C, Oby ER, Bauman MJ, Miller LE. Restoration of grasp following paralysis through brain-controlled stimulation of muscles. Nature 2012; E-pub 2012 Apr 18.
    • Hochberg LR, Serruya MD, Friehs GM, et al. Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature 2006;442(7099):164–171.
      • Ganguly K, Wallis JD, Carmena JM. Reversible large-scale reshaping of cortical networks during neuroprosthetic control. Nature Neurosci 2011;14:662–667.
        • Flint RD, Ethier C, Slutzky MW, et al. Local field potentials allow accurate decoding of muscle activity. J Neurophysiol 2012 (in press).