Published online 15 October 2008 | Nature | doi:10.1038/news.2008.1170


Monkeys move paralysed muscles with their minds

Sending brain signals through electrodes to a paralysed wrist muscle restores movement.

macaqueMacaques can learn to control paralysed limbs using re-routed brain impulses.ephotocorp / Alamy

A monkey's paralysed wrist can be moved and controlled by electrical signals artificially routed from its brain, according to scientists who say that their experiment is a step towards helping paralysed people to regain the use of their limbs.

Previously, scientists have been able to train monkeys to move robotic arms using signals routed from electrodes in their brains1. This involved decoding the activity of tens of neurons at a time to replicate actions such as grasping, and required considerable computing power.

Now, Chet Moritz and his colleagues at the University of Washington in Seattle have used similar signals to deliver direct electrical stimulation from just one neuron to a paralysed muscle.

They first implanted a number of electrodes in the motor cortex of two macaque monkeys. Each electrode picked up signals from a single neuron, and those signals routed through an external circuit to a computer. The neuronal signals controlled a cursor on a screen, and the monkeys were trained to move the cursor using only their brain activity.

The scientists then temporarily paralysed the monkeys' wrist muscles using a local anaesthetic. They re-routed the signals from the electrodes to deliver electrical stimulation to the wrist muscles, and found that the monkeys could control their previously paralysed limbs using the same brain activity. The monkeys learnt to do this in less than an hour, the team report in Nature2.

Amazing flexibility

The previous function of the neuron doesn't affect whether it can be trained to move a particular muscle. "All neurons could be used equally well, regardless of whether that neuron was originally related to the activity of these muscles. This dramatically expands the potential population of neurons that could be used to control a neural prosthesis," says Moritz.

For Andrew Schwartz, a neurobiologist at the University of Pittsburgh in Pennsylvania whose team published work in Nature earlier this year on using brain activity to control robotic arms1, the key result from the new study is the "amazing ability of these neurons to change the way they relate to the outside world."

"There's an amazing flexibility in the way that the system can learn," he says.

Learning how to control an action using neuronal activity was demonstrated by Moritz's co-author Eberhard Fetz in the 1970s. The novel aspect of this work, says Schwartz, is the way the monkeys were able to learn to use this process so flexibly, and to use the connection to activate their own muscles.

Clinical treatments may still be many years away, says Moritz. The monkeys' performance improved markedly with practice, but the long-term electrode implants needed are not yet practical for human subjects.

And moving one muscle with one neuron is all very well, but producing whole actions or coordinated movements is a much greater challenge, Schwartz cautions. "Multi-joint movement is orders of magnitude more complicated than this demonstration," he says.

But making direct connections from brain to muscle does avoid the hefty computer processing required to decode the signals that feed into robotic arms and other prostheses. This latest study relied on a battery-powered chip the size of a cellphone; in the future this will surely get smaller. "We already have electronics that are small enough to be worn in a shirt pocket, or hopefully in several years implanted under the skin like a pacemaker," says Moritz. 

  • References

    1. Velliste, M., Perel, S., Chance Spalding, M., Whitford, A. S. & Schwartz, A. B. Nature 453, 1098–1101 (2008).
    2. Moritz, C. T., Perlmutter, S. I. & Fetz, E. E. Nature doi: 10.1038/nature07418 (2008).
Commenting is now closed.