Earlier detection of diabetic retinopathy using smartphone AI

1 May 2019

Researchers at the University of Michigan Kellogg Eye Centre have shown that a smartphone which combines high-quality retinal imaging with artificial intelligence (AI) is able to detect diabetic retinopathy (DR) at an early stage.

DR is a complication of diabetes that results from high blood glucose levels damaging the back of the eye. Although it often takes several years for the condition to develop, it can cause permanent vision loss.

“The key to preventing DR-related vision loss is early detection through regular screening. We think the key to that is bringing portable, easy-to-administer, reliable retinal screening to primary care doctors’ offices and health clinics,” said Yannis Paulus, vitreoretinal surgeon and study lead author.

Scientists worked to develop a device which transformed a smartphone into a retinal camera, known as the RetinaScope. Unlike traditional retinal cameras which are expensive, large and immobile, RetinaScope is cheap and able to be used without specialist training.

The device was used to collect data from 69 adult patients with diabetes. Images of their retinas taken after induced pupillary dilation were analysed using an AI eye screening system, which categorised them as referral-warranted diabetic retinopathy (RWDR) or non-referral-warranted DR. The same images were also evaluated by two expert readers trained to recognise signs of DR in order to check the validity of the device.

The study measured whether the AI screening method were able to detect disease (sensitivity) and if it was able to confirm its absence (specificity). The AI interpretation had a sensitivity of 86.8%, significantly higher than that of an ophthalmic screening device, and specificity of 73.3%. Although one of the human image graders did achieve a significantly better level of sensitivity at 96.2%, both had far lower specificity, at 40% and 46.7%.

Researchers were encouraged by these results and are now looking into how to improve the technology, in particular attempting to find ways to use it without the need for pupillary dilation. They will also be seeking US Food and Drug Administration (FDA) approval to roll out the test commercially.



Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.