Artificial Retinas Bring Vision Back To The Blind

Patters writes "Researchers from the University of California and the Doheny Eye Institute have successfully implanted a tiny electronic eye implant with a video camera mounted on a pair of sunglasses into 6 patients, allowing them to detect light and motion. The implant is a 4-by-4 grid of electrodes which connects to damaged photoreceptors (rods and cones) on the patient's retina. It works by stimulating the photoreceptors, transmitting signals through the optic nerve to the brain. The implant only works on patients with degenerated rods and cones, and is named after Argus, the Greek god which had 100 eyes. If the implants continue to be a success, the artificial retinas could be available to the public within the next 3 years."

Comments

Obvious transhuman consequences left out

[Eunuch]

Transhumanism is like libertarianism--an obvious solution invisible to the mainstream. Hey, I don't want blind spots. I consider them to make me disabled. And to fix that I want completely artificial eyes.

Re:Obvious transhuman consequences left out

[commodoresloat]

Transhumanism is like libertarianism

You mean it's an unworkable fantasy dreamed up by conservatives who smoke pot?

Sounds about right to me.

Re:Obvious transhuman consequences left out

[fbjon]

Actually, the blind spots are an artifact of the physical construction of the human eye. It's where your nerves leave the eyeball.

Re:Obvious transhuman consequences left out

[MyLongNickName]

I think you misunerstand my post. Look at folks who have never had sight. Studies have shown that other parts of the brain begin to use the "dead space" that would have been dedicated to vison. I don't think it is unreasonale to expect that you COULD get super sight.... but something else would suffer as a result.

The brain is a very amazing creation. However, it isn;t sitting around with 90% unused capacity as is the common old wives tail. Make vision better, something else must suffer.

Re:Obvious transhuman consequences left out

[Valar]

The funny thing about libertarians is that they are dismissed by both sides of the political spectrum.

Libertarian: I think people ought to be able to do anything with their personal lives, just as long as it doesn't hurt anyone else.
Republican: OMG?! What if they smoke drugs and make gay-like?
Libertarian: Companies are just lots of people. So they ought to be able to do just about anything that doesn't hurt anyone else.
Democrat: OMG?! They'll enslave us all!!!!! (ignoring the important clause about do no evil)

Re:Obvious transhuman consequences left out

[The Only Druid]

"You might get it to work if you implanted them in an infant, but that would be kind of rough having to buy new eyes frequently during periods of rapid growth."

No, you'd never have to buy new ones: newborns arrive with eyeballs the same size as an adult. That's why children seem to have such large eyes: their skull is smaller than an adult, but they have the same size eyes. The only large-scale change to your eyes over the years is a slight shift in flexibility of different tissues (resulting in various vision issues), excluding serious degenerative issues.

Re:Obvious transhuman consequences left out

[jonadab]

> Actually, the blind spots are an artifact of the physical construction of
> the human eye. It's where your nerves leave the eyeball.

*Those* blind spots are only actually blind spots as long as you keep your gaze focused in one spot without looking around -- which you generally don't do, except when you're very sleepy, drugged, or deliberatly focusing your vision on a particular thing that's stationary (and normally when you focus your vision on one thing, it's a thing in motion, so this doesn't come into play then). Otherwise, those "blind" spots have almost no impact on your vision. As your gaze passes over something, it theoretically blinks out for an instant (if it's the right combination of small enough and far enough away), in that your retina is not perceiving it for that instant, but as your cerebral cortex processes and intereprets the informatin it is receiving, it smoothes that over automatically and fills in the blanks for you. The retinal blind spot makes a fascinating "optical illusion" science demo at places like COSI, but as long as you don't stare straight ahead like a zombie, it presents no very significant problem to your vision in practice.

However, the other poster was talking about the rather larger blind spots that result from the limits on your peripheral vision (both horizontally and vertically): in a nutshell, you don't normally see behind your head. Since your head (and whole body) don't generally move *nearly* as often as your eyes, this larger blind area has a much more significant impact on your visual perception.

4 x 4?

[blue_adept]

wow, and I thought 640 x 480 was low resolution.

Re:4 x 4?

[madaxe42]

Ah, but imagine a Beowulf cluster of these!

Re:4 x 4?

[shimmin]

I've seen a video clip of a person using the implant. He was an older man who had originally been sighted, but had lost his vision several decades ago. An object was placed in front of him on a table, and he proceeded to move his head around in a circular pattern, kind of like a bird doing some sort of mating display. I think this motion multiplied the effective resolution of the device, giving him a better sense of where the object's boundaries were than if he'd held his head still.

After about 20 seconds, he announced, it's a cup, and he was on the verge of tears, and honestly, so was I. Increasing the resolution of the device is just a matter of engineering. The concept works.

Re:Wee

[wjsteele]

It's not a dupe... it's for the OTHER eye.

Bill

remember everyone

[UlfGabe]

people blind from birth will not be able to use this to see. Their brains havn't even developed the "code" to interpret the optic nerve signals.

people who have lost eyes, or through macular degeneration, will be able to regain some of their lost visual freedom.

excellent work scientists, keep it up.

Re:remember everyone

[kebes]

You are absolutely right. Many experiments have shown that if vision is impaired during certain critical developmental periods, then normal vision will never be possible, even if their eyes work perfectly. (The work began with Hubel and Wiesel's work on kittens, for which they received the 1981 Nobel Prize in medicine, but has been extended by many others.) These experiments have even shown that you can limit vision in certain ways (blocking out only one part of a visual field, for instance, or letting them mature in an environment devoid of a particular class of visual cue) and the animal will simply have that part of their visual system undeveloped (while other parts still work).

So there is no way that those born without vision will ever attain what we consider normal vision. That having been said, it may be possible that they can achieve some rudimentary visual capabilities. For instance, they may learn to use the stimulus from a 4X4 grid in order to help them know when objects are approaching, or to better interpret their other senses. It isn't much, but for someone who has been blind their whole life, even some vague visual information (like knowing how bright their surroundings are!) may be helpful. Obviously more research is necessary in order to know if even these limited abilities can be learned later in life.

Re:remember everyone

[Anonymous Coward]

people blind from birth will not be able to use this to see. Their brains havn't even developed the "code" to interpret the optic nerve signals.

Hopefully scientists will be able to find a way of stimulating the areas of the brain to develop this "code".

Then the people blind from birth will be able to sing I can see clearly now my brain has grown.

SCNR.

Re:remember everyone

[BWJones]

OK, so like the other posters have said, there is considerable processing that occurs in the brain. However, most people are not aware of how much visual processing actually occurs in the retina. Hint: it is considerable.

As for the results that Humayan et al are showing to great effect, there are major problems aside from the engineering ones. First off, part of my PhD dissertation was on just this problem of retinal degeneration. It turns out that the implants they are designing are not taking into account some of the most basic issues of biology. Notably that any time you deafferent a CNS system, it remodels. They will have to deal with remodeling and continuously degenerating retina. In order for implants like this to work, we need to arrest retinal remodeling or take advantage of it to enable wiring into bionic or artificial biological circuits.

From an engineering standpoint, traditional electrode grids like this will end up with other problems. Notably, the issue of heating. You don't want to cook your retinas, so the need for very small currents with microelectrodes are what will be necessary. I show one such bionic implant on my blog here.

Re:remember everyone

[mikael]

Not really. Human vision develops at the early stage, at the same time as we learn to crawl along as infants. There was once a case where some parents tried "accelerate" their kids development, by skipping the crawling stage, and just using a baby bouncer instead. Apparently, the kid never learnt the concepts of "perspective" and "distance". as a consequence, she couldn't understand why objects changed in size.

There was also a guy in a 3rd world country who had cataracts since he was born. Doctors managed to help him see again, but he could only see colours, but not shapes. He still had to touch the object to get the idea of its shape.

There are so many aspects of vision that we have to learn in order to avoid becoming confused: shadows, reflections, texture, shape from shadow, perspective, not forgetting spacial relationships (partially obscured, behind, inside).

Re:remember everyone

[Dachannien]

One thing that happens during early infant brain development - when the low-level brain facilities for vision are also developing - is the mass culling of synaptic connections. That is, the brain initially grows (during late prenatal development) an abundance of synapses, far more than it needs for normal operation. A large portion of these synapses are removed during early childhood (first two years or so), with learning and experience creating a "survival of the fittest" scheme of determining which synapses to keep and which to lose.

It is almost certain that the excessive culling of synapses in the visual centers of the brain that results from not having any visual stimulus during the first two years of life is irreversible. Possibly the brain could be stimulated to produce new synapses in those areas, but it is likely that the process would cause far more harm than good.

Does it use Linux?

[snutte]

If it runs on Linux im willing to poke an eye out just to get one! :D

panning your head

[goombah99]

In principle one pixel would be enough, if you could pan your head and remember what you saw at each pixel. With 16 pixels this is simplified. Your 4096x1024 pixel scanner on your desktop does not have 4 mllion sensors, it has just 1/4000th of that number: 1024 and it uses them in a pushbroom fashion. Those 360 degree pan cameras also just use a narrow slit they push broom. Same with many sattelites.

the question is whether your brain is up to of synthesizing a image from a pan and deconvolving the large pixels down to high resolution. There's some evidence it might be able to synthesize the image from the pan since it already does that for your blind spot. And the ganglia in the eyeball do some deconvolution already so that might be possible too.

I guess we'll find out when the blind people tell us.

It would be better to grow new ones...

[mark-t]

It would be better to grow brand new biological eyes (compatible with the intended recipients DNA), and have those implanted rather than electo-mechanical solutions. One key advantage among many being that such replacements could actually grow with the person, and recipients would not be limited to adults.

Re:It would be better to grow new ones...

[cnettel]

A bit like it would be better to just teleport people than going with your typical airline?

It's hard enough to grow some kinds of human cells, and growing them in an orderly fashion to get the exactness necessary for something like normal vision, is very far away right now. I think it's quite likely that an artificial implant with a good interface will be a good-enough, or even better-than-original solution.

Alternatives...

[lxt]

...it's probably worth pointing out the research already done in various other areas - I believe a few months ago the Univeristy of Wisconsin completed a test whereby a grid of electrodes was placed on the tongue of a blind person, who wore a head mounted camera - light intensity would trigger impulses sent to the grid. Apparantly one of the subjsects even managed to navigate around a maze using it. I'm sure a /. story was posted about it...

But even this was based on previous research - I remember about similar experiments done in the late 1980s, albeit on a far lower resolution and using a extremely pad of electrodes mounted on the chest.

No news here.

[qualico]

*sigh* This story has been around for years. Here is a better resolution version from 2000: Artifical Retinas

So? It's better than nothing...take hearing:

[lxt]

It really is better than nothing - take cochlear implants. Nobody who recieves an implant (which works) complains about the quality of the sound produced...and it really is far removed from what we hear (imagine everything sounding like it was being spoken by Daleks, and you'll get the picture).

As with all technologies, you'd expect the resolution to improve over time - in the case of cochlear implants, sound quality has improved with increased numbers of electrodes being used in the cochlear, and the size of speech processors has been reduced to the point where they now look like typical hearing aids.

However, I'd imagine surgery wise, although it can be extremely complicated to insert a cochlear implant (especially if the cochlear itself is deformed), it's a hell of a lot easier to upgrade / repair a damaged implant than it would be to upgrade / repair a retinal implant.

Can you imagine being there...

[Sivar]

When someone who has been blind for their whole life sees for the first time, with a device that you and your team designed?

THIS is the true value of science.

...not totally true

[lxt]

I don't think this is totally the case.

Having some experience with cochlear implants, I can tell you children who are born completely deaf - ie, have never heard sound in their life - often adapt (over time) to cochlear implants.

However, most adults cannot do this - the brain of an infant obviously is under constant development, and so can learn how to "hear" far more easily than a totally deaf adult.

Re:...not totally true

[Anonymous Coward]

Yes, you are right. I'm a cochlear implant recipicent. I got my CI when I was 3 in 1989. I can hear and know the sounds fine.

But adults are a completely different story. They might be able to hear, but they might not be able to "learn" sound discrimination. An example of this is being able to tell the difference between low frequency and high frequency sounds necessary for speech discrimination. Of course, I was implanted early so I just ingrained that information naturally. An adult-implanted person will have a much harder time doing this. For what it's worth, I recall the auditory memory area of your brain, if you haven't heard anything in your entire life, would be remapped for other uses. Perhaps that's why it's so difficult for adults to learn sound?

I've been extremely lucky to be implanted at such an early age to develop my oral and auditory skills. Many deaf people would oppose to a cochlear implant; I didn't have a chocie, but if I did, I'd have still gotten it. It's a great thing!

AGAIN!!!

[NeedleSurfer]

These inventions seems to appear often here

But rarely in real life...

Anyone else thinking what you could do with this??

[V_drive]

How about:

- A jack that accepts video signal from a computer for work or GAMING
- Backward or otherwise mounted cameras at all times giving "rear view" (eyes in the back of your head!) appearing off to the side of the main image
- Your personal HUD! News, stock ticker, email, personal alerts and reminders, responding to voice activated commands
- Night vision or infrared
- Television receiver with subtitles
- Zooming lenses

Okay, none of that will be helpful with 4x4 res, but think of the possibilities for future use!

Then again, think of the pranks you could pull on someone by splicing it.

What is the resolution of the human eye?

[Infinite Entropy]

What would be the 'resolution' of the human eye? I think its gotta be in the range of maybe 5000*5000 to 10,000*10,000. But maybe its higher. I know I read a cool story by Greg Egan about transferring your mind into a computer and how the visual data was generated by raytracing backwards from the simulated retina, one ray per cell just like how its one ray per pixel. But I wondered at the time how many rays that was.

Re:What is the resolution of the human eye?

[elgatozorbas]

IIRC the resolution of the eye is approxmately 0.3x10^-3 radians (in both directions), based on the optics of the eye lens. The resolution of the rods and cones themselves may be lower.

Argus NOT a Greek God

[Danuvius]

Argus is a *giant*, not a God, in greek mythology.

He did have 100 eyes though. "He was thus a very effective watchman, as only a few of the eyes would sleep at a time; there were always eyes still awake.", as the Wikipedia notes

Scientific American Frontiers on PBS

[Orp]

Scientific American Frotntiers, the PBS science show hosted by Alan Alda, recently did a segment on this technology and how it worked for a man who was blinded as an adult. The other segment was on a deaf girl who received a cochlear implant.