Prosthetic limbs have gotten more lifelike — and also more useful — recently. But how do you let people feel what they're touching? Recently, scientists have developed a number of supersensitive artificial skins, but the goal of restoring sensation has remained elusive. That is, until now.
The sense of touch is incredibly important — not only does it allow us to manipulate objects, it's also a vital part of emotional communication and it gives us a sense of embodiment. Naturally, to restore the sense of touch to those who've lost it, you'd have to electrically stimulate specific portions of the primary somatosensory cortex, which is the main sensory area of the brain that deals with touch sensations.
"Over the last 15 or so years, the idea of doing this has been floating out there," said Sliman Bensmaia, who runs a somatosensory research lab at the University of Chicago.
However, there are two main hurdles to applying such a technique, Bensmaia tells io9. One challenge has to do with trying to understand the brain enough make it all work. Then there's the technological difficulty of developing electrode arrays that reliably and robustly interface with the brain. "And you have to think ahead," Bensmaia added. "For a human patient, you can't implant it and then explant it; it has to last a lifetime."
Previous animal studies have only been proofs-of-concept, which showed that restoring touch using so-called intracortical microstimulation (ICMS) is possible. But none have actually demonstrated how it can be done.
So Bensmaia was very skeptical when he was first approached to take part in the Defense Advanced Research Projects Agency's (DARPA) Revolutionizing Prosthetics project, which seeks to create an artificial upper limb to restore motor function and sensation in amputees. "But I thought it offered a lot of promise, and that we could use this as a tool to understand the brain better," Bensmaia said.
For their study, Bensmaia and his colleagues focused specifically on the sensory aspects of the limbs — they set out to identify brain patterns associated with touch, and then project sensations using ICMS. Of course, this raises the question: How does one target a specific area of the body and project sensations onto the hand, as opposed to, say, the foot?
Luckily, scientists had previously figured out that corresponding areas of the brain activate in response to touching specific body parts. In effect, the primary somatosensory cortex essentially contains a sensory map of the body. You may have even seen the image to the left (or similar images), which shows the positions of different body parts across the somatosensory cortex.
Using this information, the researchers decided to investigate three major aspects of touch: Location, pressure and timing.
To start, they trained Rhesus macaques to discriminate between different indentations, or pokes, of the hand. The training involved poking two of the primate's fingers on a hand — so if the second poke occurred on a finger to the left of the first finger poked, the macaque had to quickly look to the left.
Training complete, they then simulated the poking by stimulating neurons corresponding to the different fingers. For example, they would physically poke the index finger on the left hand, and then electrically stimulate neurons associated with the pinky on the left hand. They found that the macaques responded as if they had been physically poked (so in this example, they looked left).
Next, the researchers probed the primates' ability to discriminate between different pressures on the hand using another setup that also involved quickly looking left or right. "What happens when you poke with different forces?" Bensmaia said. "You have a greater and greater number of neurons that become activated, so one way we can mimic that or reproduce that is simply by increasing the current applied to the neurons."
After testing the primates, the team was able to create an algorithm that specifies how much current they needed to elicit the sensation of specific pressures. Again, they found that the macaques responded to the different electrical stimulations as if they had really been poked with varying pressures. They even poked a sensor on a prosthetic hand with various pressures and converted the indentations to electrical stimulations — the macaques responded as if they're own hands were poked.
"This illustrates that we were successful in relating pokes to the hand to electrical stimuli that create the same sensation of pressure," Bensmaia explained.
Finally, Bensmaia and his colleagues studied contact event sensations. When you grab an object, knowing exactly when you first touch the object and when you stopped touching it is important. In fact, the somatosensory cortex shows a huge, transient burst of activity when you first touch and stop touching an object. The researchers found they could mimic these phasic bursts (and the events they correspond to) with electrical stimulation.
The future of prosthetics
Engineers have recently made a lot of progress in creating mind-controlled robotic prosthetic limbs, but the viability of these devices is diminished if they don't include sensory capabilities. For example, how do you know if you are squeezing something too hard if you can't actually feel it? Incorporating these three aspects of sensory feedback — contact location, contact force and contact timing — could greatly enhance the functionality of prosthetic limbs.
But this work is just a start — there is a lot more to touch than what the researchers have investigated so far. "When you grasp an object, you also have information about shape, texture and whether there is movement along your skin," Bensmaia said. The team is now interested in doing more work along these lines.
Another important test will be to transplant the research to human subjects. The study shows that restoring touch is possible — at the very least — in Rhesus macaques (whose sensory systems are similar to humans'), but researchers now need to test it in humans.
If all goes well, the work could someday help both amputees and patients with spinal cord injuries. "Imagine you are tetraplegic and you have a child," Bensmaia said. "Can you image the importance of being able to touch your child for the first time?"
Check out the study in the journal PNAS.