Eyenuk Announces FDA Clearance for EyeArt Autonomous AI System for Diabetic Retinopathy Screening
On August 05, 2020, Eyenuk, Inc. announced that it has received FDA 510(k) clearance to market its EyeArt® autonomous AI System for Diabetic Retinopathy (DR), a leading cause of blindness among American adults.
Today, diabetes is the leading cause of new cases of adult blindness in the US, a problem exacerbated by the fact that there is a shortage of eye-care providers and doctors must focus on treating patients suffering from the disease rather than catching the problem early enough to start effective treatment.
DR, a common complication of diabetes, is projected to affect more than 56 million people worldwide by 2030. While DR screening is recommended for all diabetic patients, less than half get annually screened because there is a lack of eye care specialists to meet the DR screening needs. Even for those receiving their annual screening, ophthalmology appointment wait times can be weeks or more.
To address this problem, researchers and vendors have developed AI algorithms that help accurately detect diabetic retinopathy. Companies in the US, China, Portugal, and France have created detection algorithms that have produced excellent results in their own clinical trials. However, performance in real-world settings has yet to replicate these positive results.
Researchers at the Veterans Affairs Puget Sound Healthcare System and the Atlanta VA Healthcare System from 2006 to 2018 used the algorithm-based technologies on retinal images from nearly 24,000 veterans. They found the algorithms’ performance varied when analyzing images from patient populations in Seattle and Atlanta care settings, meaning the algorithms might require additional training on a wider set of images.
When it comes to the difference between human and AI detection, the former uses forms of reasoning unavailable to AI while the latter can detect pixel-level changes in tissue invisible to the human eye. As has been seen in so many studies about AI, the best systems seem to combine AI-based algorithms with human reasoning. The ultimate goal for diabetic retinopathy screening is for AI to augment, not replace, human radiologists.