To an AI, Every Eye Tells a Story

March 20, 2019 0 By JohnValbyNation


Sundar Pichai, CEO of Google


R. Kim, Chief medical officer of Aravind Eye Hospital

Ten cents won’t get you much in the American health care system—maybe a Band-Aid, if your HMO is feeling generous—but in parts of India, where nearly a quarter of the world’s blind population lives, it will cover the cost of a vision screening. Across the state of Tamil Nadu, the Aravind Eye Care System has set up a network of rural teleconsultation centers, each one supervised by a trained technician. When a patient comes in, the technician performs a basic workup, snaps photos of the inside of the eye, and sends a digital report to one of Aravind’s doctors, who calls in a diagnosis and a course of treatment. According to R. Kim, chief medical officer at Aravind’s hospital in Madurai, nearly 2,000 patients take advantage of these services every day. Yet he foresees an even breezier ophthalmic future, one powered by artificial intelligence: “You put a coin in a vending machine at the airport or the railway station, it takes pictures, and within a few seconds it tells you ‘Hey, you have this problem in your eye.’ ”

Four years ago, a joint team of researchers from Google and Aravind began work on an automated tool for detecting diabetic retinopathy, one of the leading causes of blindness worldwide. (India is home to 74 million people with diabetes.) First, they trained an algorithm to recognize the signs of the disease—distinctive spots and bleeding in the retina, the light-sensing tissue at the back of the eye. Then they began feeding it new data from Aravind’s centers. When supplied with a patient’s retinal photos, the algorithm can spit out a diagnosis in a matter of seconds. For now, Aravind’s doctors still check its work, but soon—once it receives regulatory approval—the AI will go solo. Is Kim worried about losing his job to automation? “Not really,” he says. The easier screenings get, the more patients will get screened. “I have a feeling that we’ll be put to more work when AI comes into play, because it’s going to detect many more problems,” Kim says. Similar tools could soon spot glaucoma and other vision-killing conditions.

Sundar Pichai

Silicon Valley trope that is every bit as absurd as people think:
“The ubiquity of kale. It really is everywhere. I don’t believe people who say they like it.”

Although Aravind’s rural vision centers are modestly equipped, they still require specialized gear. A retinal camera will set you back thousands of dollars, and it’s not something you’d want to lug into, say, a pop-up clinic or a refugee camp. But there may be a cheaper, more portable solution. Two years ago, a group of researchers published an article in the Indian Journal of Ophthalmology describing a remarkably effective DIY retinal camera. The ingredients: a smartphone with a plastic cover, a relatively inexpensive condensing lens, and about a dollar’s worth of PVC pipe, sandpaper, and electrical tape. Long before state-of-the-art ophthalmic vending machines begin dotting the world’s airports and railway stations, a setup like this could enable vision screenings on the fly. You take a picture, upload it to the cloud, and get your diagnosis in moments.

R. Kim

The career choice I didn’t pursue:
“Architecture. I still enjoy playing with Legos!”

Eyes have been called many things by many people—the interpreters of the mind (Cicero), the lamps of the body (Saint Matthew), the windows to the soul (anyone with a keyboard). In strictly neurological terms, though, your retinas are extensions of your central nervous system. They’re rooted in the brain, and they have all sorts of stories to tell about what’s going on beneath your skull. Earlier this year, for instance, Google debuted an algorithm that can identify a person’s sex and smoking status and predict the five-year risk of a heart attack, all on the basis of retinal imagery. (The same AI can also “infer ethnicity.”) As Kim notes, what makes these results so exciting is that the algorithm picked up on problems that the people who trained it couldn’t. “This is not something the human eye can see at this point,” he says. “There’s something beyond that the machine is seeing.” Medical researchers are actively studying the retina as an early-warning system for dementia, multiple sclerosis, Parkinson’s, Alzheimer’s, and even schizophrenia. To understand the body, look to the eye.

This article appears in the October issue. Subscribe now.

MORE FROM WIRED@25: 2013-2018

  • Editor's Letter: Tech has turned the world upside down. Who will shake up the next 25 years?
  • Opening essay by Virginia Heffernan: Things break and decay on the internet—that's a good thing
  • Edward Snowden and Malkia Cyril: Turnkey tyranny
  • Satya Nadella and Jenny Lay-Flurrie: Mindful tech
  • Susan Wojcicki and Geetha Murali: Getting girls into tech
  • Jennifer Doudna and Jiwoo Lee: Taming Crispr

Join us for a four-day celebration of our anniversary in San Francisco, October 12–15. From a robot petting zoo to provocative onstage conversations, you won't want to miss it. More information at

Related Video


25 Years of WIRED

WIRED is turning 25! We are celebrating in San Francisco this October with four days of events honoring the ideas, innovations, and icons who have shaped the world we know today—and those who will shape it for the 25 years to come.