FDA permits marketing of AI software that autonomously detects diabetic retinopathy

By Dave Muoio
Share

The FDA has granted diagnostics company IDx’s De Novo request to market its AI-based software system for the autonomous detection of diabetic retinopathy in adults who have diabetes, called IDx-DR.

This decision represents the first AI-based diagnostic system authorized by the FDA for commercialization in the US that can provide a screening decision without the need for clinician interpretation, according to the agency. The news comes just months after the FDA’s De Novo approval of Viz.AI, another AI software tool that analyzes stroke indicators and highlights CT images that could require additional clinical attention.

“Early detection of retinopathy is an important part of managing care for the millions of people with diabetes, yet many patients with diabetes are not adequately screened for diabetic retinopathy since about 50 percent of them do not see their eye doctor on a yearly basis,” Dr. Malvina Eydelman, director of the Division of Ophthalmic and Ear, Nose, and Throat Devices at the FDA's Center for Devices and Radiological Health, said in a statement. “Today’s decision permits the marketing of a novel artificial intelligence technology that can be used in a primary care doctor’s office. The FDA will continue to facilitate the availability of safe and effective digital health devices that may improve patient access to needed health care.”

IDx-DR’s algorithm analyzes images taken with the Topcon NW400 retinal camera and uploaded to a cloud server. Within minutes the software provides doctors with a binary result, either indicating that more than mild diabetic retinopathy is present and that the patient should be referred to an eye care professional, or that the screen is negative and should be repeated in 12 months. Notably, since the system’s results do not require clinician interpretation, it can be used by providers who aren’t normally involved with patients’ eye care, according to the agency.

"The FDA's authorization to market IDx-DR is a historic moment that has the potential to launch a transformation in the way U.S. healthcare is delivered," Dr. Michael Abràmoff, founder and president of IDx, said in a statement. "Autonomous AI systems have massive potential to improve healthcare productivity, lower healthcare costs, and improve accessibility and quality. As the first of its kind to be authorized for commercialization, IDx-DR provides a roadmap for the safe and responsible use of AI in medicine.”

The FDA based its decision on data from a clinical study of 900 diabetes patients’ retinal images collected from 10 primary care sites. Here, the rate at which IDx-DR was able to correctly identify more than mild diabetic retinopathy was 87.4 percent, while images with mild or lesser diabetic retinopathy were correctly identified 89.5 percent of the time.

IDx-DR is strictly autonomous, an important distinction from the AI-based clinical decision support tools that have become something of a focus for the FDA as of late.

In accordance with the 21st Century Cures Act, the agency released draft guidance in December that outlined regulated devices based on the degree of human involvement, as opposed to doing so based on risk. Reactions from industry experts and analysts were mixed at the time, but comments submitted to the FDA in February by the CDS Coalition — a group that includes representatives from clinical societies, payers, providers, software developers, and other parties — has since framed the draft as a misstep. In a recent contributed column, Epstein Becker & Green’s Bradley Merrill Thompson asserted that the agency has so far fallen short on its promise to take a risk-based approach to clinical decision support software.

“By law, FDA is supposed to focus its limited resources on higher risk technologies both to maximize public health protection, and at the same time to avoid stifling innovation where regulation is unnecessary,” Thompson wrote in the column. “Unfortunately, in developing its new approach to regulating clinical decision support software, FDA has chosen to ignore that basic principle as well as a promise the agency made to Congress and the public.”