Seeing straight with the help of artificial intelligence

The eye may be a window to the soul, but the retina is a window to health and wellness. A retinal test can help diagnose sight-threatening diseases, high blood pressure, heart disease, cancer, Alzheimer’s and other dangerous conditions.

AEYE Health, an Israeli startup with offices in Tel Aviv and New York, is on a mission to make early detection of those illnesses – starting with preventable blindness – as easy as taking a photograph.

One day in the not-too-distant future, screening for potential loss of sight “will be as easy as walking into a CVS, putting your chin on a chinrest, and learning within seconds if you’re at risk of going blind or not,” says Dr. Zack Dvey-Aharon, co-founder and CEO of AEYE. (The name sounds like “AI.”)

In the United States alone, some 35 million diabetics and 90 million prediabetics are at risk of developing diabetic retinopathy, the prime cause of blindness at working age. Another 55 million Americans aged 65 or older are at risk for glaucoma and age-related macular degeneration. Some 75 percent of the high-risk population does not undergo annual screening.

Going blind

“Millions of people are going blind just because they aren’t getting tested. This is how we will put an end to the problem,” Dvey-Aharon says.

AEYE uses artificial intelligence to capture all the data needed for a diagnosis from a single image that can be taken by any healthcare professional, nurse or pharmacy technician using a hand-held fundus camera, which captures an image of the retina. Existing systems require multiple images from a large desktop device used by specialists and cost twice as much.

It is a simple procedure, but because the existing equipment is expensive and bulky, few ophthalmologists administer the test and too many patients are unaware of its importance, unable to reach those specialists, or unwilling to cover the cost.

“Leveraging the power of AI to promote adherence to screening protocols is nothing less than sight saving,” says Shlomit Schaal, MD, PhD, MHCM, professor and chair of UMass Medical School’s Department of Ophthalmology and Visual Sciences. “Early detection and intervention are crucial in preventing severe vision loss.”

More efficient

The AEYE system is based on a proprietary algorithm that scans the image and sends an AI-based diagnosis back to the device. The whole procedure takes less than two minutes. If an abnormality is spotted, the patient can be referred to a retinal specialist – a much more efficient use of time and money.

Diabetic retinopathy is currently the only area where the FDA allows the use of AI for diagnosis independent of a doctor’s opinion. A reimbursement code recognized by health insurers was approved before AEYE Health developed its technology.

AEYE’s product is currently investigational, and the company is seeking FDA clearance for it. Results from the company’s Phase 3 clinical trial announced in February showed AEYE’s single-image software was able to capture a diagnosable reading in over 99 percent of cases and able to detect diabetic retinopathy with high accuracy even when using a smaller, portable and more affordable fundus camera. The system also showed an unusually low occurrence of false negatives or positives.

“It’s not just important to have good algorithms,” says Dvey-Aharon, an AI prodigy who began university studies on data analytics at age 12 and served in the IDF’s elite 8200 technology unit. “A large focus of our research work is to make our algorithms achieve top results on small, portable devices.”

Primary care

The only currently approved solution on the market requires a large and expensive camera generally used only by specialists, and also requires two images for each eye, which complicates the technicalities, making it much less practical for a typical primary care clinic.

AEYE Health’s technology, if and when approved by the FDA, should be significantly more accessible. The company has formed a partnership with camera manufacturer Optomed to create a camera/software bundle that doctors can subscribe to or rent for a monthly fee, instead of paying tens of thousands of dollars upfront to use a different solution. Dvey-Aharon believes this approach is better suited to the needs of the US market.

Dr. Sean Ianchulev, professor of Ophthalmology at New York Eye and Ear Infirmary of Mount Sinai and a member of AEYE’s board, sees enormous promise in the company’s technology.

Exciting times

“These are exciting times for ophthalmology and population health, as we see high efficacy of AI-based diagnostic approaches which can lead to scalable, almost exponential applications in the screening of eye disease and prevention of blindness,” Dr. Ianchulev says.

Meanwhile, in a collaboration funded by the US-Israel Binational Industrial Research and Development Foundation (BIRD), UMass Medical School, together with UMass Memorial Health Care, is conducting a pilot to show how AEYE Health’s solution can be integrated in a hospital setting.

“Screening results can talk to electronic medical records, which talk to the scheduling system,” says Dvey-Aharon. “If this model works well, we can copy it to other hospitals and that could save sight for a lot of people.”

The AI diagnostic system could also flag other diseases and send patients to specialists within the hospital environment. AEYE’s research demonstrated glaucoma diagnosis with encouraging results that could establish the company as a leader in that field as well.

Family physicians are also coming on board.

“The use of a compact retinal camera and AEYE Health software in our practice will assure that every patient can be screened during one of their routine visits each year,” says James Ledwith, MD, PI and assistant professor of Family Medicine and Community Health at UMass. “We anticipate that physicians using this technology will improve screening performance from 30-35 percent to 80 percent or more. When one out of four patients with diabetes has retinopathy, primary care screening will result in new vision-saving intervention.”

“Early screening is the difference between going blind and seeing 20/20 for another 20 years,” says Dvey-Aharon. “It’s a project with a big impact, and worth dedicating ourselves to.”

AEYE Health is currently funding on the OurCrowd investment platform. For more information, click HERE.