Could AI be a financial boost for diabetes care?

In the bowels of a stationary McDonnell Douglas-10 aircraft, parked on the runway of Bangladesh’s Shah Amanat International Airport, local eye surgeons practiced performing complex surgeries with the help of virtual reality and artificial intelligence, guided by leading experts from around the world. This is Orbis International’s flying eye hospital, a mobile training hub that deploys to lower-income countries to strengthen ophthalmology expertise. Experts said that AI, used to diagnose diabetic retinopathy, glaucoma, and macular degeneration, could reduce the financial and resource strains on local health systems. AI could, they said, be an innovative financing tool, specifically for eye health.

“We have very limited human resources. We have hardly any retina doctors at the district level. All [of them] are concentrated in the urban areas,” explained Dr. Lutful Husain, an ophthalmologist working with Orbis in Bangladesh. “So this is the way we can detect the diabetic retinopathy at the district [level] ... It reduces the suffering of the patients and also the unnecessary costs associated with referrals.”

Diabetic retinopathy is a consequence of diabetes. In Bangladesh, where 12.5% of the adult population has diabetes, this noncommunicable disease is recognized as one of the main causes of preventable blindness. While diabetic retinopathy can be treated with injections, laser therapy, or surgery, few diabetes patients go for an annual eye check, as recommended by the United States’ Centers for Disease Control and Prevention, meaning it often goes undetected. Between demonstrations inside the flying eye hospital, Husain told Devex he estimates that 25% of people with diabetes have diabetic retinopathy.