How do neural implants work?


When I Google search, all explanations I see either come up with descriptions that are too technical, or aren’t specific in how implants actually do the things. One of them actually used the phrase “in some way”.

A neuron sends an electrical pulse. An implant reads it. But how??

The implant sits on the top of the brain. How is it picking up all these millions of signals throughout the brain and interpreting it?


In: 6

It sounds like you have one thing wrong. Implants is as the name implies implanted into the brain, it does not just sit on top. It requires surgery to put the implant into the body and connect its wires to the nerves. Nerve cells sort of uses electricity through salt water to get signals from one side of the cell to the other. It is not exactly the same as an electrical wire but close enough that the nerves and the wires will exchange signals.

You would expect it to be difficult to connect the wires to the correct neurons but this actually seams like a simple problem. Just connect it to any nerves, hopefully in the right part of the brain, and the brain will quickly learn which neuron does what. We are still figuring out how to do more advanced communication but for simple things it appears to be very easy.

There are other ways of interfacing with the neurons rather then using implants. You can put electrodes on the skin and measure the electric field. As the neurons are firing they do generate electric fields in the body and these can be measured on the outside. It is very hard to receive the signals from a single neuron but works on larger signals to for example muscle groups. Signals the other way is similarly very low resolution and can basically just operate muscles.

Imagine you’ve got a giant blob of wires. The blob is giant and it’s hard to see what is attaching to what but you can see where it hooks up to the stuff outside of the blob, this screen here and this speaker there, so you know when it does stuff. You also know that when stuff happens there is a lot of electrical signals that pass through those wires.

So you think, what if I sent a probe that can detect electricity and put it into the blob. Then you would know that when the screen goes red if the wires near your probe were involved in doing that.

So what if you make a big pole with sensors all up and down it and stick it in? How about dozens or hundreds of poles?

Eventually you have a pretty good visual map of which wires are active during all sorts of different actions like when the speaker goes “baahmp” or the screen does the star fade.

So what do we do with this? Well maybe we just want to know of the wires involved in the speaker going “baahmp” are also involved when it goes “wabawabadingdong”.

Or maybe we make a giant blob that has silver wires instead of copper wires and we want to know if it works the same.

Or maybe our speaker is broken so we made a new speaker but we can’t really hook it up cause it has a different connection point. So we figure out what wires are involved when the speaker does certain things, when when those wires are active we get the new speaker to do those things.

And in some specific cases we can even send electricity through the pole and through the probe and into the wire to make the screen or speaker do things we want them to do.

This doesn’t work always though. Most of the blob is to fragile and our poles too big to stick poles into it and the probes are too bulky to only pick up signals from a single wire. But we are working in fixing those problems for the future.