Australasian Science: Australia's authority on science since 1938

The Bionic Eye Is In Sight

The Hatpack simulator. Image: Monash Vision Group

The Hatpack simulator. Image: Monash Vision Group

By Namita Bhojani

After conquering the bionic ear more than 30 years ago, Australian scientists have set their sights on the bionic eye.

In 1970, Professor Graeme Clark led the team that developed the bionic ear. He implanted it successfully into the first patient in 1978 and, since then, the cochlear implant has helped tens of thousands of hearing-impaired people. Now, more then 30 years later, the race is on to develop a bionic eye, and Australia is once more a serious contender.

Blindness affects 39 million people worldwide, so a device to restore functional sight to at least some of them promises enormous potential. Labs all over the world are developing various versions of a bionic eye, with groups in California and Germany already trialling retinal implants in patients.

The Australian government has committed $50 million through the ARC Research in Bionic Vision Science and Technology Initiative for the development of a functional bionic eye. Two proposals were approved for funding from 2010 until 2014, with Bionic Vision Australia (BVA) and Monash Vision Group tackling the bionic eye challenge from different entry points into the visual system.

The Eye
A healthy eye works by focusing an image onto the retina, which is a sheet of light-sensitive cells (photoreceptors) and neurons at the back of your eye. Neurons in the retina translate the light signals into electrical signals that are carried along the optic nerve to the brain, where they are processed further to give us a view of the world.

The bionic version of the visual system will consist of multiple components, including a camera mounted on a pair of glasses, an external processor about the size of a mobile phone, and the electronics to deliver the information into our visual processing pathways. The camera will capture images from the patients’ environment and transmit them to the data processor. The processor will then convert the image into electrical signals – the language of the brain.

The two Australian teams are using the electrical signals to target different levels of the visual processing pathway. BVA will transmit the signals to an electrode array implanted at the retina, whereas Monash Vision Group is sending the signals directly to an implant in the cortex. In both cases the system will replace the function of a healthy eye and deliver artificial visual stimuli to the visual processing pathway.

Stimulating the Retina
BVA’s multi-disciplinary team spans two states, one territory and more than five organisations. “Our idea is to implant devices into the retina that are capable of stimulating the retina with an array of electrodes,” says Dr Hamish Meffin, senior researcher at NICTA, an organisation in the BVA consortium. “We can mimic natural vision by stimulating a particular pattern across this electrode array.

“There are two devices that BVA is building – you can see them as two generations,” Meffin says. The first generation is a wide-view device that consists of 100 electrodes and aims to help patients manoeuvre around large objects. The second is a high-acuity device that has 1000 electrodes and may allow patients to recognise faces and read large print.

“We’re actually putting the two different devices in two different places,” Meffin says. “The wide-view device is sitting some distance from the retina, and it’s good because it’s surgically very straightforward. The drawback is because it’s far away, there’s a limit to the resolution you can get. [The high-acuity device] we’re implanting right inside the eye on the inside surface of the retina.”

Meffin says there are significant technical challenges with the high-acuity device compared with its wide-view counterpart, but the potential result is better.

Bypassing the Retina
The Monash Vision Group is a collaboration between Monash University, Grey Innovation, MiniFAB and Alfred Health, and they are developing an entirely different implant. “For people who have damage to the retina or have damage to the optic nerve that transfers the signal from the eye to the brain, we are essentially bypassing that damage,” explains Dr Nicholas Price from Monash University.

“A lot of people have the impression that all of your sense of vision is performed by the eye but that’s not the case. Most of vision is actually computational processing, and conscious vision happens in the brain,” Price says.

“Around 20–30% of the brain is devoted almost exclusively to visual processing. We’re targeting that first part of the brain pathway called the primary visual cortex,” Price says. “We’ll have hundreds of electrodes actually directly implanted into the cortex. A pattern of electrical stimulation comes up (from the processor) to a cable where the signal is wirelessly transmitted across the skull to the tiles of electrodes which are implanted in the primary visual cortex. We don’t need the eye. We don’t need the optic nerve. We can just directly stimulate the cortex.”

The implant is a grid of 10–20 electrode tiles. Each tile supports up to 50 electrodes, totalling 300–1000 electrodes. “The electrodes are thinner than a human hair. They barely penetrate into the brain,” Price says. “There’s no pain associated with any of this electrical stimulation [as] there’s no actual pain receptors in the brain. Essentially we’re just giving a natural form of stimulation that the brain normally receives.”

The Clinical Viewpoint
As the two groups are targeting different points in the visual processing pathway, their two versions of the bionic eye will suit different patient needs. “We’re almost targeting different patient groups, which makes us complementary in our approach,” says Dr Jeanette Pritchard, general manager of the Monash Vision Group.

BVA’s retinal implant replaces the function of damaged or lost photoreceptor cells in the retina. “We have a couple of very specific patient groups – people with retinitis pigmentosa and people with age-related macular degeneration,” Meffin says.

Retinitis pigmentosa (RP) is the primary cause of inherited blindness, and age-related macular degeneration (MD) is responsible for half of legal blindness in Australia. “In those diseases, the photoreceptors have died, but some of the other neurons in the retina remain healthy enough that we can target them with electrical stimulation.” However, BVA’s retinal implant is not appropriate for patients who do not have any healthy retinal cells at all.

In contrast, Monash Vision Group’s cortical approach could help patients who have lost retinal function altogether. “If there is damage to the optic nerve or physical trauma to the eye, like patients who have potentially lost an eye to an accident or had a crushed optic nerve, the retinal implant won’t function because it’s relying on everything to be normal from the retina back. We’re bypassing all of that and going straight to the visual cortex, so if the optic nerve is crushed it actually doesn’t matter for our technology,” Pritchard says.

However, a significant amount of visual processing occurs before electrical signals reach the visual cortex. A challenge with the cortical implant is that it skips out on this earlier processing, so it may be more difficult to recreate meaningful images.

Both teams are on track for their first patient tests in the next few years. Their first patients will be completely blind but will have previously been able to see. This is crucial because the scientists are relying on the normal visual processing pathways to interpret the artificial stimulation, but in someone who was blind from birth these pathways may never have developed. Additionally, no one yet knows how this sort of prosthetic device will work in someone who still has an element of functional vision.

As with any new technology, over time these devices will be improved and perfected. Ideally, it will be possible to update the system without further surgery. “The idea is to put most of the complexity outside, so that what we implant can be updated with the external components,” Pritchard says.

The Challenges
Even with all the new technology and ingenuity devoted to these projects, the bionic eye cannot compete with nature’s version of vision. “A lot of people think you’re going to get full 20/20 vision, [but] with an implant it really isn’t like that,” Pritchard says. “If you deliver stimulation to a single electrode then the percept that you generate is sort of a blurry white blob that is spatially localised – it’s called a phosphene,” Price explains.

“The idea with all types of bionic vision, whether they’re targeting the retina or targeting the cortex, is that by generating lots of these different phosphenes in different parts of the visual field you can essentially recreate objects. If I have three of these phosphenes in a row then I get a line. But it’s not high definition vision that people are normally used to,” he says.

For now, colour vision is also out of the picture. The photoreceptors in the retina are responsible for colour vision and, for the time being, their role cannot be replicated.

There are a lot of challenges in developing a bionic eye, such as the miniaturisation of all the electronics, transmitting power to the electrode implants and finishing with a safe, practical and portable device. When you consider that the Cochlear implant targets 22 frequencies, you get a glimpse of the enormous technological progress required for the bionic eye project. Imagine trying to see with just 22 pixels.

Researchers are using a variety of extensive experiments to test out their devices. For example, both teams have built simulators that researchers wear over their eyes to mimic what patients will see with a prosthesis. “We have staff wandering around the lab wearing these goggles, and they’re given certain tasks such as moving objects from one place to another,” Pritchard says. “That teaches us quite a lot.”

Other experiments include using jelly models of brains to practise surgical techniques, and testing the longevity of the implants. “The aim is to produce devices that will last 20-plus years, just as a cochlear implant lasts essentially the lifetime of the patient,” Meffin says.

There are similarities in the two Australian projects where the teams can collaborate regardless of their different methodologies. Both teams will need to satisfy regulatory guidelines before progressing to patient tests, and they can also join forces to recruit patients for clinical trials. Furthermore, both teams will need to devise a rehabilitation program for their patients.

“We don’t immediately expect we can turn this device on in a patient and they’ll be able to see and they can drive home. They will have to essentially learn how to see again, learn how to reinterpret visual signals,” Price says. “They might have to move their head in different ways to get motion information, just as someone who has lost an eye moves their head from side-to-side to get depth information.”

Despite all the challenges, a functional bionic eye is almost here. Both the projects have realistic timelines for a final product.

As with the Cochlear implant, it is not necessarily the first product on the market that is most successful, but instead it is the best product that wins.

Indulge for a moment and think how wonderful it would be for Australian science if we could have both the bionic ear and the bionic eye to our name. More importantly, imagine how these devices could improve the quality of life for visually impaired patients. It looks like we will see it all happen within our lifetimes.

Namita Bhojani is a freelance science writer, and an Education Officer with the CSIRO and Monash Science Centre.