Photovoltaic retinal prosthesis restores high-resolution responses | Hacker News Reader

Photovoltaic retinal prosthesis restores high-resolution responses

nature.com

113 pointsfinphil21 comments

This is a really cool approach - I covered it yesterday and here's the EPFL news post as well.

https://techcrunch.com/2021/03/18/quest-for-prosthetic-retin...

https://actu.epfl.ch/news/retinal-implants-can-give-artifici...

Still lots and lots of work to do but it obviates the issues with previous retinal implants that you basically needed to run a cable into the eye. In retrospect photovoltaics are obvious because that's more or less how the eye works in the first place.

This is one of those great ideas that seems obvious in retrospect. I hope it pans out. I was wondering how this stimulated the optic nerve, but it seems like it doesn't do so directly, but rather it interacts with the existing retina?

Yeah, this would be for blindness that's upstream from the nerve, for instance an opsin deficiency or something that makes the light-sensitive cells in the retina not work. The underlying (actually overlying since the retina is kind of inside-out) nervous network might still function fine, it just doesn't have an input. The photovoltaic dots would create a very small localized charge that would stimulate the higher-level cells rather than the rods and cones.

Well done.

ELI5 anyone? Is this an artificial retina that so many people would benefit from? If so, does it connect to an optical nerve?

In the retina, there are photoreceptor cells that take in light and emits excitatory signals to other cells in the retina that then do some basic pattern recognition and send the signal to the brain (Some of this pattern recognition is re-done in the brain but that's not really relevant in this instance).

You can excite neurons by giving them electric pulses via electrodes.

Some people are blind because their photoreceptor cells can't convert light into electric pulses (Retinitis pigmentosa). These people can be helped by artificial implants that do this conversion step, like the one proposed here. The device is split into little compartments and when light reaches one of them, the compartment converts the light into an electric signal, with an electric contact right on the cell. There seems to be no amplification involved, it's only living from the energy of the light. The tissue then gets that signal and processes it as vision. Cool, isn't it?

Disclaimer: not an expert, so correct me if I'm wrong.

Thank you! Do you know how does it connect hardware to a "wetware"? "Just plug the nerve into this socket"?

In older retinal prosthesis work it was a microelectrode array, basically dozens or hundreds of tiny electrodes piercing the retina and running a current. In this case the dots turn light into current so when you shine a light on them they activate the retinal cells near them.

Ok, it goes to existing retina, not substituting but adding some input signal gain, am I correct?

So there is still no "interface" to connect to an optical nerve directly for people who do not have functional retina...

No, no direct optic nerve interface. That isn't feasible with any technology we know of. The nerve bundle, like the spinal cord, is incredibly dense withe fibers. There's also pre-processing and network stuff that goes on in the retinal layers that you don't want to skip.

Fortunately (though I may be wrong here) I believe a lot of blindness upstream from the optic nerve is capable of being addressed by artificial lenses, corneas, replacement fluid, retina augmentation etc. The nerve and central nervous system really represents where it starts getting beyond our depth.

As someone with progressive retinal tearing this is great news. I've been reading about methods of turning interneurons into photoactive elements using simple chemicals injected into the eye but this seems a lot more straightforward.

Sufferer of RP here. The brain really is amazing and capable of a lot of inference on sparse data. Unfortunately it’s inference is not always correct, or intuitive. See Charles Bonnet Syndrome.

I am curious how they could train to combat it.

This is amazing! I have not finished reading it, but I wonder whether they are already thinking about human augmentation. I have average vision, but would love to be able to see infrared and UV thanks to an implant.

I love the idea, but there may be privacy implications. It reminded me of learning about cameras that can accidentally see through clothes that manufacturer had to to recall to remove.

You already can (if you are willing to be a clinical trial guinea pig). There are nanoparticle eye drops that enables that.

Could you elaborate or give some links?

An article about it working in mice: https://newatlas.com/nanoparticle-near-infrared-vision/58686...

Assume that they are now doing human trials.

Probably worth mentioning that although the headline of that article refers to "eye drops", the sources that it cites say that the testing was actually done with an injection into the retina itself.

We are still a good way off replicating the resolution of the human eye in an implant, and generally we have a hard time matching the impressive dynamic range of our eyes. Give it another 20 years.

Until then something that converts infrared and UV to the visible range is a lot more reachable (whether that's an implant like an intraocular lens, simple contact lenses, or just eye drops)

As someone who stands a decent chance of developing age related macular degeneration 30 years from now, I love any and all news like this.

Very cool. Wonder if this will be a killer app for the upcoming Apple VR glasses.

Reply on news.ycombinator.com