A Scientist Says Human Tissue Can Compute Complex Equations

In the world of computing, many breakthroughs and innovations mention ways to make the silicon brains in our machines more like the biological brains in our bodies. In fact, there’s a whole field dedicated to the concept: neuromorphic computing. But as the line between man and machine continues to blur, could the same be said in reverse? Are there ways that our bodies—specifically, our living tissues—could develop computational abilities?

The researcher behind a new study published in journal IEEE Access investigated this possibility with a framework known as “reservoir computing.” The idea is that data can be fed into complex “reservoirs” capable of encoding rich patterns. By analyzing human tissue, Yo Kobayashi—the lone author on the paper—discovered that the soft biological structures possessed many of the properties that would make them particularly adept at this type of computing.

“Common reservoirs include nonlinear dynamical systems like electrical circuits or tanks of fluid,” Kobayashi said in a press statement. “There are comparatively few studies that use living organisms as reservoirs, and until now, none that use in vivo human tissue.”

So, how does one transform a biological system into a computational one? Kobayashi’s technique involved gathering biomechanical data by asking participants in the study to bend their wrists at various angles. Then, Kobayashi took various ultrasound images to capture all of the miniscule muscle deformations in the wrist, which allowed him to construct a “biophysical reservoir” for data processing.

“An ideal reservoir possesses both complexity and memory,” explains Kobayashi. “Since the mechanical responses of soft tissue inherently demonstrate stress—strain nonlinearity and viscoelasticity, muscular tissue easily satisfies these criteria.”

Being viscoelastic means that this biophysical reservoir is “viscous” (like a liquid) and elastic, which allows it to retain a physical memory of its past deformations. Kobayashi benchmarked his breakthrough by testing the human tissue using complicated nonlinear equations, and the model using the biophysical reservoir turned out to be more accurate than a model that used standard linear regression. If this type of computing can be improved, it could a big deal for a variety of technologies.

“One potential application area of this technology is wearable devices,” Kobayashi said in a press statement. “In the future, it may be possible to use our own tissue as a convenient computational resource. Since soft tissue is present throughout the body, a wearable device could delegate calculations to the tissue, enhancing performance.”

So, could our bodies one day be computational devices? The question may not be as strange as you might think.

You Might Also Like