April 19


AI Device Gets FDA Approval and Takes Detection of Diabetic Retinopathy to The Next Level

By Gabriela

April 19, 2018

Diabetic retinopathy is a diabetes-related complication which impacts the eyes. This develops due to damage to the blood vessels of the light-sensitive tissue found at the end of the eye, that is the retina.

In the beginning, this complication might cause only mild eyesight issues or have no signs at all. In the worst-case scenario, it can lead to blindness.This condition might occur in those who have type 2 diabetes or type 1 diabetes.

The less controlled the blood glucose and the longer the diabetes diagnosis, the more likely to have this complication. Therefore it is essential to prevent it and watch for some early signs. And the last approval from FDA might help with that.

AI device gets FDA approval and changes things.

AI Device Gets FDA Approval

United States health regulators announced that it would approve the first medical machine which uses AI, i.e., artificial intelligence software in order to discover more than a mild level of the most prevalent cause of loss of vision in 30 million Americans with diabetes.

This device is known by the name IDx-Dr, and it is manufactured by low-based IDx LLC. Also, it is the first AI device to get FDA authorization which offers a screening decision without needing a clinician to read the results or image.

That means that this device can be used by healthcare providers who are not usually part of eye care, like primary care physicians who are in more frequent communication and contact with those who have diabetes.

According to IDx, the goal is for the providers to screen patients without the need for an eye specialist to interpret images for signs of diabetic retinopathy.

More About the Device

This device was reviewed under the new regulation of the Food and Drug Administration i.e., FDA, made to speed to certain market machines seen as low to moderate risk and machines without the prior legally marketed device.

This is part of the efforts of the Commissioner Scott Gottlieb to streamline approvals on different fronts such as a cheaper version of usually expensive biotech drugs and generic drugs.

Malvina Eydelman, who oversees the division of the agency of ophthalmic, and also throat, nose and ear device talks on this topic.

According to Malvina, the Food and Drug Administration will continue to ease the availability of effective and safe digital health machines which might improve the accessibility of the patient to the necessary health care.

This device will be used to discover diabetic retinopathy.

How Does This Device Work?

The program of this device uses artificial intelligence software to analyze a picture of the eye taken with the help of a retinal camera. These images are captured by Topcon NW400 camera and will be uploaded to a cloud server; this is where the software resides.

What this software does is that it shows the physician that the patient either has more than just mild diabetic retinopathy and needs to be referred to an eye care professional or is negative when it comes to more than mild retinopathy and needs do another screening in one year.

Based on a multi-center trial of this recently-approved device in up to 900 of those with diabetes there is a sensitivity of 89.5 percent and 87.3 percent.

According to Eydelman many people with the condition aren’t suitably screened for diabetic retinopathy. It is like that since 50 % of these people don’t visit their eye doctor every year.

So in order to manage this, early detection is necessary, and the newly-approved device can help with that.


  • Gabriela

    Co-Founder & Editor-in-Chief of Diabetes Health Page, Fitness trainer and instructor has dedicated her career to educating and informing people for over 10 years. As one of the most passionate diabetes advocates, Gabi has worked tirelessly to ensure that those people receive the education and support they need to properly manage their diabetes and achieve their health, fitness and weight loss goals.

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}