Slashdot Mirror
Bionic Eye Patient Tests Planned For 2013
angry tapir writes "Australian researchers are getting ready to test a bionic eye on patients in 2013. The eye consists of 98 electrodes that stimulate nerve cells in the retina, which is a tissue lining the back of the eye that converts light into electrical impulses necessary for sight, and allow users to better differentiate between light and dark. With the bionic eye, images taken by a camera are processed in an external unit, such as a smartphone, then relayed to the implant's chip. This stimulates the retina by sending electric signals along the optic nerve into the brain where they are decoded as vision."
...right down the tubes if your bionic eye suddenly decides to start humming Bjork tunes and your Google phone joins in...
Operation Guillotine is in effect.
Please, don't quote the line. You know the one. The one with three comparatives. It's too predictable.
One of the challenges I see is that the optical nerve, isn't really a peripheral nerve (connecting peripheral sensors to the central nervous system), but something connecting 2 parts of the central nervous system. Beside other peculiarities stemming from this, it has a result which makes the bionic eye much more complicated than other organ replacements:
The signal is already processed. Light get detected in the deeper layer of the retina (where the cones and rods lives), transmitted to the upper layer (nerves cells doing this transmission plays the same role as peripheral nerves) and gets processed in the upper layer.
The optical nerve doesn't carry simply levels detected from the cones and rods, instead it carry some shape information (boundary detection done by comparing signals from neighbouring groups of cones and rods) and colour contrast information (done by comparing the signal of a small group of cones with surrounding cones). (The same kind of pre-processing going into the spine or the crianial nerve's nuclei).
A bionic eye will need to similarly pre-process the image, and then manage to send the correct output to the correct type of fiber.
On the other hand, the various later stages of the visual pathway in the brain do further processing on the signal (line detection, shape detection, motion detection, etc...), so the brain might manage to make something useful out of the signal even if it isn't optimal at that stage.
I wonder how functional and useful the resulting perceived image would be for the patient. Well, probably better than nothing, but still...
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
hmm thats interesting, I'm interested to see how that turns out, helping people with sight problems. Somewhat clever idea too.
What next?
I hope they will have a nice visor that goes over my bionic eyes, that lets me see infrared, warp core plasma, and all sorts of exotic radiation. I promise to sing the "Reading Rainbow" song while wearing it.
sudo make me a sandwich
So the image processing is done on a smartphone, then sent to a chip in the eye, presumably wireless.
If it is being processed on an external device it is likely that the camera will not be in the eye (at least at first), and it appears that any camera is supported which leads to some interesting possibilities (streaming TV or the internet direct to your optic nerve anyone?) and also some interesting hacking opportunities. To bad that installing something like this would require you to lose an eye however it could lead the way to space opera style cyborgs.
I assume that 98 electrodes means a resolution similar to 98 pixels so it sadly wouldn't provide a very good replacement, however this will probably improve in time as historically eye problems have attracted strong support and funding. It would also not work very well for people with damaged optic nerves and would probably require the removal of a natural eye if the patient has one.
null
Can someone explain how sending signals to the brain stimulates the retina? Do they bounce and come back out or something?
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
I never asked for this
I seem to recall reading an article about this 10-15 years ago, in which they had an 8x8 array of sensors that were directly stimulating areas in the visual cortext. Fully blind patients were able to correctly identify a number of shapes.
However, if I recall correctly, the nerves' responsiveness to the unnatural direct electrical stimulation wore off quite quickly, and the assessment at the end was that the electrical/neuronal coupling was going to be the main problem to overcome.
Do they get to call this a "Retina Display" then?
Just think of the possibilities when the eye is given the ability to see beyond the optical wavelengths.
Faith is a willingness to accept something w/o complete proof and to act on it. Reason allows you to correct that faith.
You're absolutely right. Your bullshit superficial analysis based on very few facts definitely shows that first bionic eye won't be very useful.
Sorry, but where again did I say that no bionic eye will ever be possible?
I said there's an interesting challenge. And probably the first generations of bionic eye will have a hard time replacing the whole retina including this processing functionality. In fact, apparently due to this exact challenge, the first planned bionic eye don't replace the whole retina, but stimulate the middle layer (the one transmitting the signal from receptors to the upper processing layers), thus not replacing the whole retina, but only the photoreceptor (still useful for some disease, like macular degeneration as reported elsewhere in this thread, but not yet as useful for other disease like diabetes where the whole retina dies)
where uninformed opinion is rated Informative
It happens, not only that IAAMD, but that I did my bachelor-level thesis (well equivalent thereof. It was before the Bologna treaty in europe and the splitting of university cursus into bachelor and master) on bionic implants. Got to interview scientist working on such implants (research teams working bionic eyes, and surgical team using cochlear implant). This challenge is exactly what said bionic eye researcher told us too. And given the above, this is also considered for the upcoming bionic eyes.
But yeah, its just easy to troll around and bash people making point about interesting challenges arising in some research.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
There are a couple of groups working on commercial bionic eyes, one implant with 1,500 electrodes (that's actually high enough resolution to be useful for recognizing objects) did allow some previously blind people to read text, in a human study between 2005-2010:
Link to the original paper with PDF download: HTML, PDF
Published in the journal Proc Royal Soc B: http://royalsociety.org/news/retinal-implant/
In the media: forbes.com February 2012
That's new to me, can you give a citation where I can learn more detail about this?
Most of the Neurophysiology books I've read during my medial studies. I'll have to pick up a specific reference.
Meanwhile Wikipedia isn't that bad and has some explanation of how the signals are processed.
In college I took a physics class concerning light and optics, and the professor said that seeing isn't a function of the eye, but of the brain. Perhaps it was a matter of his being a physicist and not a biologist.
It's a good enough approximation for a light & optics course.
It's just that once you go into the tiny details, you might need to be more precise.
"seeing isn't a function of the eye, but of the brain."
is somewhat correct, except that the first steps of this brain processing is already taking place in some layers of the retina, because it has more in common with the central nervous system than other sensory organs.
"Seeing isn't a function of the optical eye, but a function of the neural processing done on said signal"
But then you'll need to go into details about the various neural circuitry, which well beyond the scope of a *physics* course.
Just like newtonian physics are a good enough model for my field of work and I don't need to think about Einstein or quantum mechanics or string theory or whatever.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
I guess instead of the 6 million dollar man, it would probably be more like the 6 billion dollar man, no?
This appears to be the company doing the research, they're based in Victoria but the UNSW logo appears at the end of the video:
http://bionicvision.org.au/eye
(They also say they're on track for starting human tests in 2013)
They should let Stevie Wonder be the first guy to get a shot at this shit, can't we give Stevie just a peek? Wrote songs in the key of love, just a quick peek?
I wonder what happened to the Dobelle Eye... same style but wired directly into the visual cortex: http://www.wired.com/wired/archive/10.09/vision_pr.html
[RIAA] says its concern is artists. That's true, in just the sense that a cattle rancher is concerned about its cattle.
Why stick with ones in the eye sockets? If you're already doing the processing offsite, why not have the lens remote and movable, too? It would be pretty trippy if your POV was your cellphone, always seeing yourself in the 3rd person viewpoint. Then again, the opportunities for upskirt shots or seeing around corners would be pretty cool.
I'd hold out for the IXian model.
Siri, what am I looking at?
This is an old problem. Getting electrodes on individual neurons is the holy grail of neural implants, but if you make the electrodes that small, they damage the nerves over time IImpedances around tiny electrodes are high, current densities are high, and the current needed to spred to and stimulate the nerves tends to be high enough to cause electrolysis. Electolysis is *BAD* in tissue.
When you make the electrodes large enough to reduce the impedance to something reasonable, and reduce the current density, the electrodes are too big. This is especially true in the retina, where electrodes are basically half floating in a salty bath of vitrious humor. It works well in cochlear implants because the electrodes are surrounded by a bony channel that restricts current to the local nerve,a nd the nerves are aligned in a curved line: the further in the electrode, the higher frequency the sound perceived.
With the human eye, there are multiple *classes* of sensory cell, different pigments, with several layers of sophisticated processing sells behind them doing "center surround" based processing, edge detection, and the electodes have *no way* to localize the current to the appropriate layers of cells. It's like trying to type with boxing gloves on.
How do I know about this? I *built* the highest frequency neural stimulators in the world several decades ago.