8
$\begingroup$

I'm writing a cyberpunk book where people can regain their vision by connecting an artifical eye and the optic nerve would be replaced by a electrode, is it possible? Ignoring the rejection and possible infections, could an electrode transmit information as good and as fast as an optic nerve?

$\endgroup$
6
  • 1
    $\begingroup$ @BMF The only fault I see with the Q is that I believe it should be hard-science instead. $\endgroup$ Commented Jul 18, 2022 at 19:19
  • $\begingroup$ It's worth looking up the brain's anatomy to get the picture. It's not as simple as the optic nerve goes from the eye to the brain. To make a good question read How to Ask, and check out our tour and read the help center for our guidelines. $\endgroup$ Commented Jul 18, 2022 at 19:20
  • 5
    $\begingroup$ Artificial retinas exist and even started to be commercialized. So you can electrically mimic the electrical impulses that go to the optical nerve. This was done by implanting a little chip where the rods and cones would be. The problems were mainly low resolution, avoiding heating up the eye and issues with how to wear the camera. $\endgroup$
    – UVphoton
    Commented Jul 18, 2022 at 19:22
  • $\begingroup$ @Mindwin Yeah, I think you're right. This could make a really good [hard-science] Q. I think it may need to be worded a little differently, maybe "can electrodes supplement optical nerves?" or something else. $\endgroup$
    – BMF
    Commented Jul 18, 2022 at 19:25
  • 1
    $\begingroup$ When saying "the optic nerve would be replaced by a electrode" are you aware that "the optic nerve" is actually a huge bundle of nerves? So "a (single) electrode" would not be enough. Basically a nerve is an electric conductor, but the connection point to the next nerve is neuro-chemical, so typically you would add an electrode into an existing nerve. $\endgroup$
    – U. Windl
    Commented Jul 19, 2022 at 8:59

3 Answers 3

14
$\begingroup$

Yes

Now for the ugly details.

You say "electrode." However, an "electrode" is anything used to conduct electricity. That can be a superconducting chunk of niobium to a better-than-average-grade of sand to graphite to a wonderful bit of gold. Why is this important?

  • Because a superconductor has perfect conductivity and is therefore a LOT faster than optic nerve transmission—but the temperatures required make it useless for cyberpunk.

  • Because a semiconductor (like silicon or germanium) are only as fast or slower than the optic nerve.

  • Because poor conductors, like graphite and tungsten, will pass electricity but only after generating a boat load of heat.

  • Because the obvious choice is a metal conductor that will always be faster. Except...

And then there are details you don't want to deal with.

Termination

Connecting an artificial anything to the human body is non-trivial. Connecting something that must convey information is even worse. The connection between your electrode and the optic nerve will have problems. In the electronics world, we call this termination.

Whenever you try to connect one electrical conducting object to another electrical conducting object, you have the problem of termination (the need to match the impedance1 of the two conductors to ensure maximum power transfer and minimum corruption of the information). Think of it this way. If you have a two-inch pipe carrying water at 20PSI and simply connect it to a one-inch pipe. What happens when the water hits it? All kinds of cool things! And electrically, it's even worse. You can actually degrade the signal you're trying to transmit because you failed to terminate correctly.

Corrosion

Another problem is corrosion. In the real world, when two different metals are brought together you risk what's called galvanic corrosion. In other words, the electrical properties of the two metals actually act to cause corrosion. Now, we could solve this problem in the biological world (where you have base goos and acidic goos and other kinds of goo that simply don't want the metal anywhere near it) by using surgical stainless steel, which would resist corrosion.

But the conductivity of surgical steel is about 1/40th of the conductivity of copper. It simply stinks as an electrical conductor.

Insulation

Another problem, and it's not a trivial problem, is that there isn't just one optic nerve. What we call the "optic nerve" is actually a bundle of nerves. A LOT of nerves. And that means your conductor must have as many wires as there are nerves—and those wires must be insulated one from another or all kinds of messy things like crosstalk (the coupling of energy from one transmission line to another in a way that corrupts the information on either line) and shorts (where the user suddenly sees the optical equivalent of a blue screen of death because all the nerves are being fed the same information). Insulating wires that thin is no small thing—but that's the beauty of fiction!

Magnetism

And I'll add one more even though it's a bit of a stretch (OK, a long stretch) to believe it's a significant problem. You don't want to use an electrode that's magnetic. Most good conductors aren't magnetic. But it's worth noting that the human body is remarkably resistant to magnetism. That electrode might not be. Neither might the artificial eye. One would hate to see the outcome of a knife fight to which your opponent brought not a gun, but a magnet. I'm just sayin'.

Recommendation: Avoid the pain of including too many details

Yes

You can believably use an "electrode" to connect an artificial eye to the optic nerve in a cyberpunk setting.


1You might have already heard of resistance. A superconductor has zero resistance to the flow of electricity. Resistance impedes the flow of electricity, usually by virtue of converting some of it to heat. Did you catch that word, "impedes?" Resistance is passive. "Reactance" is active (capacitance and inductance). The two together (resistance + reactance) are called "impedance," referring to all possible electrical mechanisms that can slow the march of electrons. Don't worry about this. Just know that termination isn't as simple as I'm describing in this answer.

$\endgroup$
10
  • 1
    $\begingroup$ At the speeds and distances that normal neurons run at, pretty sure you can skip the transmission line terminations... 😉 not a lot of cross-talk or reflections at a couple kHz (going off typical rise times of 0.15 - 5ms), at least, not unless you've got a few kilometers of cable. Now, if the eye had USB... well... $\endgroup$
    – Samwise
    Commented Jul 19, 2022 at 3:05
  • 4
    $\begingroup$ @Samwise I'm just pointing out issues that would be clever issues to weave into a story. But, for the record, I've dealt with crosstalk designing ABT and Multibyte circuits with frequencies in the KHz range. You're correct that the problem is rise time... but nothing stops the circuitry from rising a heckuva lot faster than KHz frequencies require. We dealt with rise times in the picosecond range despite the ability to operate in KHz ranges due to min/max loading requirements. $\endgroup$
    – JBH
    Commented Jul 19, 2022 at 5:27
  • 3
    $\begingroup$ "Because a superconductor has perfect conductivity and is therefore a LOT faster than optic nerve transmission" electric current in super conductors isn't much faster than in normal conductors. $\endgroup$
    – Bersan
    Commented Jul 19, 2022 at 8:23
  • $\begingroup$ @Bersan A great deal of this answer has been simplified to help a user who isn't an electrical engineer. When you bundle a lot of wires together, you end up with reactance (principally capacitance) between the wires. It's obviously not a lot when dealing with something as small as this, but there are a lot of parallel wires, so it's not trivial. The resistance of the wire and the reactance act to slow the propagation of a signal (look up "RC time constant" to see what I mean). The zero or near-zero resistance of a superconductor nullifies that propagation delay. $\endgroup$
    – JBH
    Commented Jul 19, 2022 at 10:22
  • 1
    $\begingroup$ Surely the best choice of metal for the electrodes would be gold? Resistivity only slightly worse than copper, and corrosion resistance is superb. Weight shouldn't be an issue (it's small) - and I suspect that even raw material cost won't be too bad. $\endgroup$ Commented Jul 20, 2022 at 15:25
9
$\begingroup$

With 2020's tech, it is actually around 8 times as fast.


DISCLAIMER

We start with a huge caveat. We do not have technology to convert digital into neural signals with the level of reliability required for video input in a resolution/depth of color compatible with the human eye, as @UVphoton points in his comment.


I don't think the transmission rate will affect it much. What will take time is the video encoding / decoding.

The system would work like this: A camera embedded in the cybereye would (1) capture video, (2) encode it, (3) send through a short cable to the back of the head (where the visual cortex is located), (4) then decode and (5) feed the signal to the neurons.

Everything but the end of step 5 is hardware-based. The brain receives signals as if the cyber-eye hardware was the optic nerve/eye.

Given the speed of electrical impulses in a cable/optic fiber is close to C and the distance is way shorter than a foot, we'll consider step (3) to be instantaneous.

As for the other steps, (ref)

Capture Post-Processing
(e.g., Bayer filter, chroma resampling) A few lines (e.g. 8) < 0.50ms

Video Compression
(e.g. Motion-JPEG, MPEG-1/2/4 or H.264 with single-pass bitrate regulation) 8 lines for conversion from raster scan A few thousand pixels on the encoder pipeline 0.25ms << 0.10ms

Network Processing
(e.g. RTP/UDP/IP encapsulation) A few Kbytes < 0.01ms

Video Decompression (JPEG, MPEG-1/2/4, or H.264) 8 lines for conversion from raster scan A few thousand of pixels on the decoder pipeline 0.25ms << 0.10ms

Display Pre-Processing (e.g. Scaling, Chroma Resampling) A few lines (e.g. 8) < 0.50ms

(The decode stream buffer step was removed because we're not buffering at the "client").

Since the neural interface does not exist, we'll assume it takes the same to convert from neural signals (which are electrical impulses) than it takes to put the signal to a video display.

EDIT:
Maybe compression/decompression doesn't make the perception faster or is necessary for sight, but it also slows the process down negligibly. And with the proper compression algorithm the cyber-eye can also do livestreams or record video into a SSD sub-cranial storage or something like that.

Now, the brain takes 13ms to understand an image from the meaty human eye (ref). Our cybereye takes about 1.5ms to transmit the data, which is more than eight times faster than the old meatbag accessory.

$\endgroup$
6
  • $\begingroup$ And you expect the brain to do JPEG, MPEG-1/2/4, or H.264 video decompression? $\endgroup$
    – U. Windl
    Commented Jul 19, 2022 at 9:02
  • $\begingroup$ @U.Windl uhm no? If you actually read the answer you're commenting on you'd quite quickly grasp that that is in fact not OPs intention and the compression is only applied to the optic fiber in between the camera and the DECOMPRESSOR in the brain (that feeds raw data to the surrounding neurons) $\endgroup$
    – Hobbamok
    Commented Jul 19, 2022 at 9:59
  • 1
    $\begingroup$ can't we skip the compression/decompression altogether and just transmit raw data? $\endgroup$
    – Christian
    Commented Jul 19, 2022 at 10:06
  • 1
    $\begingroup$ So the "decompressor in the brain" is another electronic device (allowing serial data transfer)? And on the speed: How does the decompressor in brain make optic perception faster? $\endgroup$
    – U. Windl
    Commented Jul 19, 2022 at 10:10
  • $\begingroup$ There is no need to use any complex package protocol at all ("RTP/UDP/IP etc."). Ignoring the about 150 million receptors in the retina, there are only about 1 million nerve fibers (-> en.wikipedia.org/wiki/Retina ), which gives us an upper limit of how many "pixels" we'd need (assuming one of these fibers can be represented as a 24/32 bit number). This is half of the number of pixels in Full HD, easily transmitted uncompressed in digital or analog serial connections. A tiny copper wire, or optical if you want to go fancy and stay with the theme, would suffice easily. $\endgroup$
    – AnoE
    Commented Jul 19, 2022 at 14:32
8
$\begingroup$

Currently existing product in category bionic eye.

Argus II retinal prosthesis system. Has approval of FDA.

It has helped some people restore some visual perception.

Is it full vision replacement? Very much no. Current tech has too few sensors an article mentioned trying to increase it to 150 sensors where million is the target.

Could such technologies eventually be? Probably.

Compactness and energy are the issues, not speed.

$\endgroup$
3
  • $\begingroup$ 150 "sensors" compared to 576-megapixels? $\endgroup$
    – U. Windl
    Commented Jul 19, 2022 at 9:06
  • $\begingroup$ @U.Windl Did you read the replies to that question? No, the human eye does not have that high a resolution. In color you have less than 10MP. Fortunately the bulk of your resolution is in the center of vision. $\endgroup$
    – Corey
    Commented Jul 19, 2022 at 12:54
  • $\begingroup$ At the moment, it's not "150 vs 576M", it's "150 vs 0". Somebody using one of these is still going to be legally blind - but it's a lot better than nothing. $\endgroup$ Commented Jul 20, 2022 at 15:29

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .