In December 2011, a horrific car accident knocked Jason Esterhuizen unconscious. When he woke up in a hospital in Pretoria, South Africa, hours away from his hometown, he couldn’t see. The crash had destroyed his eyes and left him completely blind.
Esterhuizen was devastated. At the time, he was 23 and studying to become an airline pilot. The first two years after the accident were the hardest. “Life changes in an instant,” he tells OneZero. “I used to fly airplanes and ride motorcycles and drive my own car.”
Esterhuizen eventually got mobility training and learned how to read braille, use assistive devices, and work on a computer. Then, in 2013, he tuned in to a TV news segment about a company working on a brain implant that could create artificial vision for people like him. Second Sight, based in Sylmar, California, had just gotten approval in the United States to sell a retinal implant designed to assist people with blindness caused by a rare genetic disorder called retinitis pigmentosa. Esterhuizen wasn’t a candidate for that device, but by 2018 the company had developed a brain implant that could change his life.
Now part of a small clinical trial, Esterhuizen is one of six blind patients to receive the experimental device, called the Orion. It’s meant to provide artificial vision to people who have gone blind from a wide range of causes, including glaucoma, diabetic retinopathy, optic nerve injury or disease, and eye injury. If it works and is proven safe, it and other brain implants could potentially help many more people who are blind.
When Esterhuizen learned he was a candidate for the trial at the beginning of 2018, he and his wife uprooted their lives in South Africa and moved to Los Angeles to be in the study. The device, he says, allowed him to see his birthday candles for the first time in more than seven years.
From the outside, the Orion looks like a pair of sunglasses with a small camera and video processing unit (VPU) attached to it. Implanted in the brain, however, is a postage stamp-sized chip containing 60 electrodes that sits on the visual cortex, the part of the brain that processes visual information. When the device is switched on, the camera captures a person’s surroundings, and the wireless VPU converts those images into electrical pulses using an algorithm. Those pulses are transmitted to the electrodes on the brain, which interprets them as visual clues.
The brain implant is what sets the Orion apart from the Argus II and other so-called “bionic eyes” on the market. Those devices don’t require brain surgery: Instead, the electrodes are embedded behind the eye. Retinal implants require recipients to have some functioning cells present in the eyes, which is why they are currently only approved for patients with retinitis pigmentosa, which affects only a small percentage of the 3.4 million Americans who are legally blind or visually impaired.
With the current system we’re testing, you don’t even need to have eyes for the device to work.
Retinitis pigmentosa causes photoreceptors, the light-sensing cells responsible for sight, to gradually die, causing vision loss over time. But other eye cells called ganglions, which talk directly to the brain, remain intact. Retinal implants, like the Argus II and those made by France’s Pixium Vision and Germany’s Retina Implant, are designed to stimulate these cells, which transmit visual information along the optic nerve to the brain. About 350 retinitis pigmentosa patients worldwide have received the Argus II device.
But the Orion, which shares much of its technology with the Argus II, bypasses the eye and optic nerve completely. “With the current system we’re testing, you don’t even need to have eyes for the device to work,” says Dr. Nader Pouratian, the neurosurgeon at Ronald Reagan UCLA Medical Center who implanted Esterhuizen’s device. As the primary investigator of the trial at UCLA, he has outfitted four patients with the device. The other two study participants received the implant from Dr. Daniel Yoshor at Baylor College of Medicine in Houston, Texas.
Despite the risk of infection or bleeding or the possibility that the implant wouldn’t work, Esterhuizen didn’t hesitate to undergo brain surgery.
When investigators first switched the implant on after the procedure, they needed to test each of the 60 electrodes one by one to see how much electric current each needed to receive before patients started seeing light. They used that information to make a custom program for each patient, a process that took months.
Now, Esterhuizen and the other participants have regained a limited amount of vision after being completely blind for years or decades. While they don’t see color, shapes, or clear edges and can’t yet read text, they are able to distinguish light from dark, they can recognize moving objects, and they have some degree of depth perception. People and objects appear as dots of light corresponding to where they’re located, and as they get closer, more dots appear. “It’s like learning a new language,” Esterhuizen says. “You learn how to interpret what’s going on.”
Patients meet regularly with vision researchers at UCLA and Baylor to test the device and learn how to use it. In one exercise, they look at a black computer screen and point to a white square that appears intermittently in different locations. The majority of the time, they can successfully point to the square.
It’s not natural vision, but Pouratian says that the device lets the participants do everyday tasks they weren’t able to before. “It’s not that the system helps them become completely independent, but if you can’t see anything, being able to see just a little bit becomes extremely valuable,” he says.
Esterhuizen says he feels safer leaving his apartment alone because he can now see when cars are approaching. Now, he can sort laundry and even find certain objects in his home.
Vision starts in the eye, but it’s the brain that recognizes images and interprets them — a process that mostly remains a mystery. Scientists know that the brain contains maps of the visual field, and that every location in that field is represented by a unique location in the brain. But they haven’t figured out where those exact locations are yet. This is why the Orion and retinal implants can only create a limited range of vision for now.
“Our electrodes are big compared to neurons, so we’re stimulating a lot of neurons at once and the brain is interpreting that,” says Jessy Dorn, vice president of clinical and scientific affairs at Second Sight. “We’re not at the level of each individual cell.”
As a safety precaution, the implant that Esterhuizen and the other patients received only uses one electrode array to stimulate the left side of the brain. As a result, they can only perceive visual cues from their right-side field of vision. If the one array proves to be safe, eventually, Second Sight plans to implant one on both sides of the brain. Dorn says the company is also working on ways to improve the technology to enhance the resolution and range of vision for patients.
The company wants to expand the number of implanted electrodes to between 150 and 200. And it’s working on improving its camera and VPU, potentially with thermal vision and facial and object recognition.
The researchers are also trying to achieve more natural vision by figuring out how to deliver the electrical stimulation in a way that better imitates the firing of groups of neurons. “This idea that we can stimulate the brain to produce visual perceptions is well-known, but the way we need to do it in order to maximize visual perceptions is not as well understood as we want it to be,” says Pouratian.
Dr. Abdhish Bhavsar, director of the Retina Center in Minneapolis and a clinical spokesperson for the American Academy of Ophthalmology who is not involved in the Orion study, says much more research on brain mapping will be needed to provide patients with a greater range of useful vision. “We have a long way to go before we understand what stimulating the brain will do in terms of vision,” he says. “If we developed a map of the brain that showed what exact parts generate what type of images or perceptions of the visual world, then we could start making models based on that.”
The early results from the Orion are encouraging, but brain implants for vision are still very much in their early days. Safety is a major concern: One patient in the Orion trial experienced a seizure after the device was implanted. People who receive such implants will need to be followed for years to make sure there are no complications that emerge later on. Electrodes in the brain also cause scar tissue to form over time, making them stop working, so it isn’t clear how long these implants will last. Second Sight’s Dorn says the electrodes used in the Orion device should work for at least five years. That means patients will probably eventually lose what little vision they acquire with the devices.
Another major limitation of the Orion is that it’s only useful for those who were born sighted and later lost their vision. In people who are born blind, the parts of the brain that are responsible for sight are not fully developed, and visual information cannot be effectively transmitted to the brain. A device that could help all people with blindness is still a long way off.
And if the FDA eventually approves the Orion, not everyone who’s eligible to get the implant will want to undergo brain surgery. The device is also likely be expensive. The Argus II retinal implant costs about $150,000, though it is covered by Medicare in many states.
Esterhuizen though is hopeful about the future of assistive technologies for the blind and visually impaired. “It’s just baby steps for now,” he says. “But eventually I think this technology will change the lives of millions of people.”