Slashdot Mirror
Scientists Create a Way For People With Amputations To Feel Their Prosthetics (gizmodo.com)
An anonymous reader quotes a report from Gizmodo: Prosthetic hands have gotten increasingly sophisticated. Many can recreate the complex shape and detail of joints and fingers, while powered prostheses allow for independent, willful movement. But a new study published Wednesday in Science Translational Medicine offers a potential glimpse into the future of the technology: Artificial hands that actually feel like a living limb as they move. The researchers recruited people with amputations who had been given surgery that reconfigured certain muscle and sensory nerves surrounding the amputated limb, allowing them to control their prosthesis through intuitive brain signals (thoughts) sent to the repurposed nerves. Across a series of experiments involving three of these patients, the researchers attached devices that generated vibrations along specific muscles near the amputation site. When the device was turned on, these vibrations created an illusionary sense of kinesthesia -- an awareness of conscious self-movement -- in the prosthetic hand as the person performed tasks with it, both in a virtual stimulation and in the real world. The volunteers had amputations that extended just past their elbow as well as their whole arm. Not only did the experiment let them "feel" their hand as they opened and closed it, but the restored intuition allowed them to perform tasks without needing to constantly look at their hand. And coupled with vision, it gave them overall better motor control over their prosthesis.
Permanent neural electrodes present a lot of problems. One is that, to trigger a nerve, enough charge has to be deposited to trigger enough neurons, reliably enough, for the sensation to be detected and learned as valid. That's worked well in stable structures like the cochlear, where the bony channel surrounding the electrodes localizes the charge, but hasn't worked so well for othe permanent electrodes. If the electrodes are in muscle, the relevant wires tend to be bulky and interfere with movement, or they're fragile and they break. Also, the smaller the actual stimulator electrode (which is typically platinum melted to amke a ball on the end of the wire), the greater the resistance (which means more electrical voltage is needed) and the higher the current density around the electrode (which can cause electrolysis and tissue damage) and the greater the electrical noise around the electrode (which obscures the signal). Conversely, you can use larger electrodes, but the signal spreads out and triggers more neurons. And with neurons embedded in normal muscular tissue, well, things move and break your wires. It's easer in the head because you can embed the wires in stable bone, and even leave a physical jack screwed down on the skull.
You can play games to localize the signals: you can use short, triggered bi-phasic pulses to keep the pulses localized. You can even make "ghost" electrodes by triggering pulses at two adjacent electrodes to focus the trigger between those electrodes, but it's tricky. The field is filled with lots of "big ideas", actually getting them to work has turned out much more difficult.
Sadly, there is not currently any good way to really localize the signals to provide good resolution. Where the nerves are already laid out conveniently, such as in the spiral of the cochlea where deeper electrodes are lower frequency sounds. For re-routing touch sensations..... It's an interesting problem. I'd never expect it to approach the sensitivity of human touch. But proprioception, feeling movement... yeah, I could see that. It has enough duration that the a low current signal can accumulate and trigger local nerves.
I suppose a clever hacker might have you feel someone else's prosthetic.