On 11 April the US Food and Drug Administration (FDA) announced the marketing approval of the first medical device to use artificial intelligence (AI) to detect more than a mild level of diabetic retinopathy, the most common cause of vision loss among diabetic patients and the leading cause of vision impairment among the US working-age population.

IDx-DR, the device developed by IDx LLC, is able to produce screening decisions without the need for clinician interpretation of retinal images, allowing the device to be used outside specialist centres, such as by primary care physicians.

Not only will this device provide the opportunity to increase the screening rate of diabetic eye disease, which is currently underdiagnosed in the US, but will also reduce the burden to diabetic patients who are presently plagued by multiple, frequent visits to varying doctors and specialists.

This approval decision paves the wave for similar AI devices to enter the market. Dr. Malvina Eydelman, the director of the Division of Ophthalmic, and Ear, Nose and Throat Devices at the FDA’s Center for Devices and Radiological Health, noted in a press release that the agency pledges to “continue to facilitate the availability of safe and effective digital health devices that may improve patient access to needed healthcare.”

In a clinical study involving retinal images from more than 900 diabetic patients, IDx-DR was able to correctly diagnose more than mild diabetic retinopathy 87.4% of the time, and was able to correctly diagnose negative for more than mild disease in 89.5% of cases.

Early detection of this disease is fundamental for managing the care of diabetes patients and preventing further vision loss. However, despite this, 50% of diabetic patients fail to see an eye specialist once per year. This device has the potential to dramatically increase the screening rate in these patients, and may help limit both the detrimental impact diabetic retinopathy can have on a patient’s life and the burden that the disease places on the US healthcare system.

This is not the first incident where the potential of AI within diabetes has been realised. An AI network developed by Cardiogram was able to detect individuals at risk of diabetes in a recent study with 85% accuracy, solely through the use of a smartwatch. Clearly, the scope and impact of AI are yet to be comprehended, not only in diabetes but also in how we view and approach healthcare in general.

IDx-DR was granted breakthrough device designation by the FDA, allowing the company to receive intensive interaction and guidance from the FDA during its development, and was reviewed under the De Novo premarket review pathway, a regulatory pathway created by the FDA for novel low-to-moderate-risk devices in spaces where there is no prior, legally marketed device.