FDA Permits Marketing of AI Device

FDA https://www.fda.gov now permits the marketing of an artificial intelligence-based device to detect a mild level of diabetic retinopathy in adults who have diabetes. The device called IDx-DR is a software program that uses an artificial intelligence algorithm to analyze images of the eye taken with a retinal camera called the Topcon NW400.

“Early detection of retinopathy is important part of managing care for millions with diabetes. However, many patients with diabetes are not adequately screened for diabetic retinopathy since about 50 percent do not see their eye doctor on a yearly basis,” reports Malvina Eyedelman M.D Director of the Division of Ophthalmic and Ear, Nose, and Throat Devices at FDA’s Center for Devices and Radiological Health.

With FDA’s decision, artificial intelligence technology can be used in a primary care doctor’s office. IDx-DR is the first device authorized for marketing that provides a screening decision without the need for a clinician to also interpret the image or results. This makes it usable by providers who may not normally be involved in eye care.

A doctor uploads the digital images of the patient’s retinas to a cloud server on which IDx-DR software is installed. If the images are of sufficient quality, the software makes it possible for the doctor to detect mild diabetic retinopathy in patients and then refer the patient to an eye care professional.

The doctor many find that the image is negative for more than mild diabetic retinopathy and the doctor can then plan to rescreen in 12 months. If a positive result is detected, patients will be directed to see an eye care provider for further diagnostic evaluations and possible treatment as soon as possible.

FDA evaluated data from a clinical study of retinal images obtained from 900 patients with diabetes at ten primary care sites. The study was designed to evaluate how often the IDx-DR could accurately detect patients with more than mild retinopathy which resulted in 87.4 percent of the time and the device was also able to correctly identify patients who did not have more than mild diabetic retinopathy 89.5 percent of the time.

Go to https://www.fda.gov/aboutFDA/CentersOffices/OrganizationCharts/ucm347835.htm for more information on FDA medical devices.

Share Button