Paralyzed woman grips, sips coffee with robot arm

View the video

Directing a robotic arm with her thoughts, a paralyzed woman named Cathy can pick up a bottle of coffee and sip it through a straw, a simple task that she hasn’t done on her own for nearly 15 years. The technology that brought about the feat is a brain-computer interface system: A computer decodes signals from a tiny chip implanted in the woman’s brain, translating her thoughts into actions that are carried out by the robot arm.

The seemingly mundane task of bringing a drink to one’s mouth is the first published demonstration that severely paralyzed people can conduct directed movements in three-dimensional space using a brain-controlled robotic device. This latest application of the system, called BrainGate, is described in the May 17 Nature.  

“Much has been demonstrated in terms of laboratory work and monkeys, but this is the first time showing something that’s going to be useful for patients,” says neuroscientist Andrew Jackson, of Newcastle University in England. A commentary by Jackson on the new developments appears in the same issue of Nature.

There’s still a lot of work to do before BrainGate can be used outside a lab. In the current design, the tiny sensor that sits in the patient’s brain is attached to a mini fridge–sized computer via ungainly wires. So making the system wireless is one goal. The researchers hope that within a decade the BrainGate system will be available and affordable for people who are paralyzed or have prosthetic limbs. Eventually, similar technology might restore function to a natural limb that no longer works.

“In many ways the real dream of the research is one day to connect brain to limb,” says neurologist Leigh Hochberg  of Brown University in Providence, R.I., Massachusetts General Hospital and the Department of Veterans Affairs. Hochberg is leading the BrainGate trials.

In BrainGate’s current arrangement, a baby aspirin–size silicon sensor is implanted in the patient’s motor cortex. In this part of the brain, neurons are laid out in a map that corresponds to parts of the body. The sensor, which contains 96 electrodes, is implanted a few millimeters deep in the motor cortex’s “arm” area. Thin wires lead from the sensor to a penny-sized pedestal beneath the skin that connects to a plug atop the head. Wiring from the plug goes to a computer. When Cathy thinks about moving her paralyzed arm, elaborate algorithms decode the signals picked up by the electrodes and translate them into commands that are carried out by the robotic arm.

The movements are still pretty clunky, says neuroscientist John Kalaska of the University of Montreal. And the new demonstration is a big step forward from using thoughts to move a cursor on a computer screen, BrainGate’s big news in 2006 (SN: 7/2/2011, p.26).

Since then, the algorithms that decode the brain signals have gotten a lot better, says neuroscientist John Donoghue, also of Brown and the original developer of BrainGate.

The research team is still figuring out how much functionality should be incorporated into the robotic limbs. “The brain basically says, ‘OK hand, go out and reach for something,’” says Donoghue. Future work will improve the robotic limbs so they fill in more of the details of a task the same way the nerves in an arm do.

---

Back Story – MINDS CONTROL MACHINES

Efforts to bridge the brain and electronics to create better prosthetics began decades ago. 

---

A paralyzed woman’s thoughts are translated into a robotic arm’s actions by a brain-computer interface system.
Credit: Donoghue et al.