We humans are melding with machines. This is not about using machines to amplify our knowledge (the Internet) but actually embedding chips within our bodies, to amplify our thoughts. The cochlear implant, used to augment hearing, is perhaps the best known brain-computer interface. But the last decade has seen great leaps, translating even memories and thoughts into digital code. Now scientists out of Tel-Aviv University have managed to build an artificial cerebellum to restore brain function in injured rats.
The cerebellum is a small part of the mammalian brain, tucked under the two cerebral hemispheres. It plays an important role in motor function, language and the regulation of emotions.
This new artificial cerebellum, created in the lab of psychologist Matti Mintz, can receive inputs from the brain stem, which filters the incoming neural information from the rest of the body.
Nearly all the work in brain-computer interface (so far) has focused on mapping the neural patterns involved in motor function. The neural correlate for motor function is relatively easier to understand than say, the neural patterns for sight or speech.
Mintz and his team followed and analyzed neural signals arriving at the cerebellum, and what signals left the cerebellum and turned into visible actions, learning or behavior. The researchers then developed a substitute circuit on a chip that mirrored the function of the cerebellum. They then wired this chip to a rat’s brain via electrodes.
In the lab they disabled the rat’s cerebellum with anesthetic before attempting to train the rat to have conditioned response: When the rat hears a tone, a puff of air will hit its eye, and so with time it should learn to blink to avoid the puff of air. The anesthetized rat never learned the connection between tone and puff of air. But when the scientists hooked the rat up to the synthetic cerebellum, it learned the connection just like any healthy animal, and closed its eye when it heard the tone.
The ultimate goal of course is to restore function to injured or diseased brains using such replacement digital parts.
Next up is study larger areas of the cerebellum and into the future, other more complicated brain regions.
There are two great challenges for the industry moving forward. One is building better chips that are more easily and firmly integrated into the brain, either just beneath the skull or embedded into the gray matter itself. The second is a bigger challenge, perhaps an impossible challenge for the industry: Finding a way, in the more complicated areas of the brain, to distinguish between the important signal and noise created by somewhat random neural firing.
Of course there is also the problem of our unique personal past experience creating a one-of-a-kind tapestry of connections between the billions of our neurons. Dissecting this sort of complexity in order to translate personal thoughts and desires into computer language remains, at least in the practical scientist’s mind, insurmountable.
Yet there are others who believe that implanting fake memories (which has already happened in lab rats) or replacing a blind man’s sight, will become reality by the end of the century. Fascinating times ahead folks.