Stay on Top of Enterprise Technology Trends
Get updates impacting your industry from our GigaOm Research Community
In 2012, a paralyzed woman with a 96-electrode sensor the size of a baby aspirin implanted onto the surface of her brain was able to think about steering a robotic arm toward a canister with a straw in it, move the canister toward her mouth, tilt it so the straw fell into her mouth, and take a sip.
It was the first time in 15 years that the woman, who was one of two patients doing experiments with the BrainGate implant, had picked up anything, let alone served herself a beverage, and the look on her face when she finished said it all.
“There was a moment of true joy, true happiness,” John Donoghue, the neuroscientist who pioneered the BrainGate implant a decade ago, said in a Brown University video. “It was beyond the fact that it was an accomplishment … it was really a moment where we helped somebody do something that they had wished to do for many years.”
It was a seminal moment in the relatively new world of brain-computer interfaces, where neuroscientists and engineers are working on building implants and robotic limbs and even entire exoskeletons to allow people whose movements are limited by spinal cord injuries, or Parkinson’s disease, or strokes to override these physical limitations and have their brains communicate directly with machinery instead.
Slowly but surely
Now more than 10 years into the BrainGate project, Donoghue’s team has moved from working with monkeys who used their thoughts alone to move cursors and play primitive video games to enabling actual people to engage with their surroundings, and with considerable success. (The 53-year-old woman known as S3 touched her intended target within a set time nearly 50 percent of the time using one robotic arm and 70 percent of the time using another, while a second patient known as T2 produced nearly identical results.)
Lee Miller, a neuroscience professor at Northwestern University, has been enjoying some success as well with his project: implanting tiny multielectrode arrays in monkeys to help them grasp, lift, and drop a ball. These arrays are programmed to detect the activity of just 100 neurons to read the signals that lead to hand movements, even though as many as a million are involved.
“We are capturing a very impoverished amount of the motor ability humans are capable of,” he said, “so even though we could get the monkey to reach out and grasp the ball with reasonable dexterity, there is no way the monkey would be playing the piano.”
Miller’s research builds on similar work out of the University of Washington that, back in 2008, used an algorithm based on the activity of just 12 neurons to get monkeys to bypass paralysis by controlling their muscles using these “artificial” connections, as bioengineer Eberhard Fetz told me.
“The brain controls normal muscles and the brain computer interface simply picks up these control signals in the brain and activates artificial muscles or artificial limbs,” Fetz said. “So instead of controlling with our fingers, we control it with the cells that control our fingers.”
Turning thoughts into movement
Miguel Nicolelis, a neuroscience professor at Duke, is hailed by many as one of the true pioneers in the field of brain-computer interfaces. He was the first to both propose and demonstrate that animals (including humans) can control prosthetic devices via brain-machine interfaces, and his development of what are called chronic, multisite, multielectrode recordings is behind scientists’ ability today to measure the activity and interactions of large populations of single neurons in the brain.
Nicolelis’ Walk Again project was on one of the world’s biggest stages this past summer when a 29-year-old paraplegic in a metal vest and a blue cap dotted with electrodes kicked off the World Cup in Nicolelis’ native Brazil. The man used his own thoughts to control the neuroscientist’s exoskeleton.
Nicolelis told CNN at the time that the kick was meant to “shock the world,” and added: “[One day] we’ll be walking in New York and we’ll see a person walking on the streets that could not walk before. I think in our lifetime we’ll see that.”
These are still the early days. The sheer volume of neurons in the brain is so large, and the complexity so mind boggling, that researchers in the fields of neuroscience and neurorehabilitation are celebrating every baby step forward, including the simple grasping and lifting of objects – and so are the participants themselves, for whom these actions represent life-changing advances.
“I remember once we were discussing the project with several potential participants,” Donoghue of BrainGate told The New York Times in 2010. “‘Would you like to walk again?’ someone asked an interested candidate. ‘No, I’d just like to be able to scratch my own nose,’ he answered.”