University of Chicago scientists aim to develop prostheses with sense of touch
Research in field languished for years — until now
Nicho Hatsopoulos is among the University of Chicago researchers studying rhesus monkeys to one day develop a prosthesis that can feel. The many cameras are used for 3D imaging in neural motor research. (Chris Walker/Chicago Tribune)
In an underground laboratory at the University of Chicago, neuroscientist Sliman Bensmaia peered at a computer attached by wires to a rhesus monkey's brain.
A lab technician grazed the animal's finger using a metal probe, and the computer screen erupted in red.
"That's pretty cool," Bensmaia said, grinning. "You can see the brain becoming active just by tapping the hand."
Next, instead of physically tapping the animal's hand, the technician planned to run a small current of electricity through electrodes in the animal's brain to simulate the probe. If the animal looked in a certain direction, the scientists would know the "virtual touch" worked.
This research is part of a groundbreaking quest to accomplish what was once the stuff of science fiction — build a machine that helps humans feel.
Funded by the Defense Advanced Research Projects Agency and spurred by the return of injured Iraq and Afghanistan war veterans, the research aims to design prostheses that will not only be able to move, but will also provide amputees and quadriplegics with a sense of touch.
Scientists have known for more than a century that applying electricity to neurons can elicit certain reactions — a muscle twitch, a sudden feeling of euphoria, a long-forgotten memory recalled. But stimulating those cells to help people overcome certain disabilities has been done only more recently, spearheaded in the 1960s by the development of the cochlear implant for hearing.
Unlike hearing or vision studies, however, touch research languished for decades, impeded by the expensive machinery needed to perform experiments and a certain "not as sexy" quality, Bensmaia said
"People take (their sense of touch) for granted more than vision or hearing," he said.
But then hundreds of wounded veterans began returning to the United States without arms or legs or the use of their limbs because of spinal cord injuries, and interest in developing better prostheses spiked. Through the DARPA project, scientists at the Johns Hopkins University Applied Physics Laboratory last year completed a new prosthetic arm that can rotate, twist and bend in 26 ways. Scientists also recently outfitted patients with brain electrodes that allowed them to move simpler robotic arms with their thoughts.
Without any tactile feedback, however, the usefulness of the prostheses is limited. Lacking the sense of touch, patients could not, for example, differentiate between corduroy and silk, a pen and a pencil or a poke and a punch.
More important, "they have to constantly be visually monitoring what they are doing or they wouldn't know whether they were holding or crushing something," Bensmaia said.
So last year, Johns Hopkins gave Bensmaia's lab about $1.5 million of its federal money to develop even more advanced prostheses that will eventually give the users a simulated sense of touch through the machine's metal and motors.
But how do you replicate the feeling of a coffee cup in your hand or the difference between a five- and a 50-pound weight? The U. of C. scientists set out to identify and replicate the qualities of touch, including texture, shape and force, through complex mathematical equations known as algorithms.
Scientists implanted platinum alloy electrode arrays, each the size of a pencil eraser, into the monkeys' brains. The scientists then created neural impulses by emitting small but focused electrical currents, and recorded the animal's behavior in response.
After simulating thousands of touch sensations, Bensmaia and his team hope to build algorithms, essentially mapping out the way the brain reads those touches. They will then use those sensory algorithms to build software for the robotic arm's computerized sensors that will transmit impulses to electrodes in the human brain, mimicking touch.
Josh Berg, Bensmaia's study director, took a step back from the U. of C. testing room and grasped at an apt summary.
"Up here, we are not vision, touch or smell," Berg said, grabbing his head. "We are all electricity. What we are trying to do is translate information into a language the brain can understand."
Since 2006, DARPA, which is part of the Department of Defense, has poured $129 million into its Revolutionizing Prosthetics program. Johns Hopkins University and its collaborators expect to implant electrodes in the first human this summer. A second patient would get implants in 2012 that would include a feedback loop, providing a sense of touch based on algorithms developed in Bensmaia's lab. And a third patient would get implants in 2013 that may allow the patient to wirelessly operate two prosthetic arms, according to Johns Hopkins researchers.
U. of C. neuroscientist Nicho Hatsopoulos recently applied to work on the development of that wireless system. Hatsopoulos, who specializes in the neuroscience of movement, co-founded Cyberkinetics Neurotechnology Systems, which was one of the first companies to implant electrodes in humans in order to control machines with their thoughts.
"Where we are right now is basically the beginning stages of 'The Six Million Dollar Man,'" Hatsopoulos said, referring to the fictional 1970s TV series about an astronaut turned bionic man. As he spoke, a rhesus monkey moved a cursor around a computer using only his thoughts.
But some worry that the technology might also be used for more ethically complicated purposes. They imagine soldiers using their thoughts to fly airplanes and maneuver combat robots in war zones. They foresee ordinary people using implanted electrodes to quickly expand their memories, download new information or augment their skills.
"It is a little scary," Bensmaia said. "It may change the world completely."
For the moment, however, Bensmaia's focus remains on the groundwork: stimulating neurons and recording those effects. In his lab, Bensmaia leaned over another scientist, attempting to isolate one neuron out of some 100 billion in a rhesus monkey's brain.
The scientist grazed the animal's third finger. Almost immediately, a thin red line spiked across the computer monitor and a rack of speakers crackled, emitting the amplified sound of a single neuron firing.
"It's a thing of beauty," Bensmaia said.