Sensing bionic limbs are here and they work

New research from Johns Hopkins University and DARPA shows how far sensing bionic limbs have come, proving that the technology is well on its way to offering real limb replacement. The breakthrough comes by way of patient interaction as much as advanced engineering, as work with real amputees shows how natural bionics sensing can really be.

The researchers recount an episode in which they decided to trick the participant by stimulating two fingers, rather than just one — the patient immediately asked if someone was playing a trick on him, by changing the rules of the test. That convinced them that the sensing is both genuine and, more importantly, natural-feeling. The patient had responded to the novel input very quickly, without stopping to interpret the meaning of the signal in the brain.

Double bionic arms, designed and installed by Johns Hopkins. Sensing is quickly becoming a limiting factor in bionic control. Fidelity in collecting control information from the brain or elsewhere in the nervous system is important — without that you can’t control the arm itself — but a limb can only be so accurate so far without feedback information about the results of the movements it makes.

Classically, this refers to the fragile cup scenario, in which a person with a bionic hand must grasp a cup of water firmly enough that it doesn’t fall, but gently enough that it doesn’t break. The only way to do this is to sense how much pressure is being applied. And, crucially, the only way to interpret how much pressure is too much is to relay this information to the brain of a human who can judge the strength of things.

Currently available information is slightly vague about where the brain is being stimulated to produce these touch sensations. But since the volunteers were able to correctly identify the finger being touched with near-100% accuracy in the very first trial, it’s likely that the electrodes are stimulating the sections of motor cortex already associated with finger sensation. This sets the technology apart from current sensing technology, which forces the patient to form new associations between totally new neural activity and familiar sensations. Other teams using a similar approach have achieved stunning preliminary results.

Of course, having passing this sort of conceptual threshold, it won’t be long before researchers start to improve the numbers involved — number that apply to natural human perception as well. Sensation coming from a bionic source does not have to be speed-limited by the diffusion of ions in solution, as are sensory neurons, or temperature-limited by the safety constraints of flesh. Bionic sensation could plausibly let a person put their hand down on a frying pan to test its temperature — and to judge it with their brain, the same as they would any reasonable level of heat.
We’re developing the hardware necessary to restore the relationship between the brain and the outside world — and in the process developing the hardware necessary to completely change that relationship forever. If your brain is wired up and you’re thousands of miles away on business, why not let your partner run your spare hand over their face, letting you literally feel a little bit of home? And you don’t have to be an amputee to get electrodes put on your brain, which opens up the area of extra mechanical limbs.
DARPA is funding this research because of the incredible potential it has to improve the lives of thousands of wounded veterans — but there’s also plenty of emergent military and industrial value to be had, in any project of this type. We live in an age when data is both money and power, and this genuinely altruistic medical research is slowly turning thought into data.