Researchers simulate what "bionic sight" may look like

2 pictures

Fast-moving objects may appear invisible to people with bionic vision(Credit: Ione Fine and Geoffrey Boynton / University of Washington)

View gallery - 2 images

It's easy to imagine bionic sight as crystal clear and even enhanced, like the augmented body parts in science fiction. But the reality could be very, very different for a typical bionic eye recipient. Researchers at the University of Washington developed visual simulations that indicate what the world might look like to people with retinal implants. The resulting images are, in a word, blurry.

The researchers simulated the vision that results from two different types of bionic eye that are currently in development. Both simulations indicate that objects will appear fuzzy, with blurred outlines or comet-like tails. If something is moving quickly, a person with a bionic eye may not see it at all.

In one of the bionic eye types considered, this is the result of a kind of input delay (or "lag," for the gamers reading). In normal vision, light first hits the photoreceptor cells (rods and cones), which convert it into electrical signals that go through multiple layers of processing, separately, on the way to the brain. But thanks to a feat of genetic engineering, in the bionic eye, incoming light waves trigger different cells than usual in the initial step – cells that aren't usually light-sensitive. These "fake" photoreceptors are much slower than real ones at responding to visual input.

The other bionic eye technology simulated in the study sends an electrical signal to the retina that excites all cells simultaneously. The problem here is that there's too much stimulation. The extra noise in the neural signals being sent to the brain causes streaks to appear in the image, while the stimulation of both ON and OFF pathways (rather than one or the other) confuses the brain and results in a weird edge effect around each object.

It had previously been estimated that bionic sight would appear somewhat like the giant scoreboards at sports stadiums – a grid of black-bordered dots forming a mostly-complete image. But according to this study, the actual results will be far less impressive.

That doesn't negate currently in-development bionic eye technology. But it does provide a great big reality check. The surgery required to get one of these devices implanted is long, invasive, and costly. For some people it won't make a difference to their decision – any vision restoration is worth the price of admission, no matter how small. For others, it might tip the scales against surgery. And in any case, more information will only serve to temper expectations.

The people most likely to benefit from the in-development devices are those who have the bulk of their vision neurons still intact. This includes people with macular degeneration, which is incurable and affects your central vision, and retinitis pigmentosa, which affects peripheral and night vision.

Bionic eye transplant devices are already beginning to hit the market. One, the Argus II Retinal Prosthesis System, was earlier this year implanted in an 80 year old man with age-related macular degeneration. Two years ago a version of the same device received US market approval for treatment of retinitis pigmentosa.

Study lead author Ione Fine notes that several more devices will be available in the next five to 10 years, and in order to improve them scientists must not only compare the strengths and weaknesses of all the different types but also develop better models of what people actually see.

"Until we do that," she says, "we're just shooting in the dark in trying to improve these implants." She also notes that her team's research is not meant to be considered a model or genuine prediction, but rather an approximate simulation and a demonstration of what more useful information a fully developed and validated model might provide.

You can see a video of one of the simulations below.

A paper describing the study was published in the journal Philosophical Transactions B.

View gallery - 2 images

Top stories

Recommended for you

Latest in Biology

Editors Choice