Caveat: not a neuroscientist, but have worked with many and been involved in MRI and fMRI imaging for investigating HCI concepts/ products and marketing. As wel

Caveat: not a neuroscientist, but have worked with many and been involved in MRI and fMRI imaging for investigating HCI concepts/ products and marketing. As well, when I was younger had trigeminal neuralgia, considered the most painful disease (explained below), and became fascinated with neuroscience. The approach of closing the loop on agonist/ antagonist feedback in muscle fibers makes an enormous amount of sense and is brilliant. Without the two “outbound requests” happening in a closed system the brain is left to approximate the difference between the slight signal shift. Even nanosecond differences in “send/ receive” requests would be registered as pain. I had a neurological condition called trigeminal neuralgia where the myelin wears down between the positive/ negative nerve bundles in the trigeminal nerve and occasionally touch. The reason it’s considered the most painful affliction is the trigeminal nerve is where your whole body’s pain, pressure, and heat sensory nerves converge. So when the +/- touch, every fiber of you feels like it’s being stabbed, burned, and crushed at the same time. About 90% of folks with it have have one occlusion in the myelin and will feel excruciating pain in a localized are. I had 9 full occlusions and 20+ partial. So lucky me, I got to experience it through my whole body. Basically, when the +/- nerves touch, no part of your body is receiving/ returning the expected “latency” for pain, pressure, or heat, and perceives the differences of return times as nerves (including those in agonist/ antagonist muscle) being split/ damaged, thus a pain response. Totally makes sense that closing the proper agonist/ antagonist fibers to create a proper feedback loop could eliminate the unexpected diff on signal requests that trigger phantom pain. EDIT: Just to give yourself some real world examples.. Do something like build a fence with your friends, all day. One person operates the saw, one the hammer, so on and so fourth. At the end of the day using the same tool in repetition, hold your tool, everyone applies blindfolds, then have an “experimenter” touch the subjects forearm at different points. They’ll experience the touch at a different point than if they weren’t holding the tool. When you use something for a long time you integrate it into your “body schema” - basically your brains map of yourself. My absolute favorite example of body schema extension are experienced crane operators in docks. They can see a container on a ship, need to place it on a massive pile they’ve been building all day, and can no longer see in their line of vision where they’ll place it. But they can still do it without being able to see where they’re dropping it. All that’s to say - your brain builds better and better models of it’s container (body) and environment by developing more and more precise models of input/ output diff expectations and essentially a “path prediction” algorithm. When the actual data is outside of parameter/ estimates get out of whack, we experience things like pain where there’s no harm. But in an amputee, agonist/ antagonist muscle fibers are no longer on the same circuit. The body adapts, and sends two different “ping” requests to both channels. So I guess to simplify all that: When theres a micron of difference in length in key paired muscle fibers (as is common in amputations, injury, etc), the brain sends two signals, diffs the return times and if they’re out of wack with past lived experience you get pain. While the amputee still doesn’t have the “normal” limb length the body is expecting, closing the loop provides the expected send/ receive timing variance.