Eyenuk receives FDA clearance for additional camera for autonomous detection of diabetic retinopathy

The FDA clearance paves the way for the EyeArt system to receive an upgraded image quality assessment module and Eyenuk’s proprietary Real-Time Image Quality Feedback solution. (Image courtesy of Adobe Stock)

Eyenuk received FDA clearance to use an additional retinal camera for the detection of diabetic retinopathy (DR) that is easier to use and make it easier for primary care clinics to adopt the technology.

According to a news release, the company can now use the Topcon NW400 retinal camera with its EyeArt AI system to automatically detect diabetic retinopathy (DR), adding to the already-cleared usage with Canon CR-2 AF and Canon CR-2 Plus AF cameras.

The company noted the regulatory clearance of the EyeArt v2.2.0 system makes it the first and only AI system that is FDA-cleared for use with multiple retinal cameras by different manufacturers.

The FDA clearance builds upon the EyeArt system’s debut nearly three years ago as the first and only AI technology to have FDA clearance to detect both more-than-mild and vision-threatening DR. In the European Union, the EyeArt system is the first and only AI system approved under MDR Class IIb to detect DR, age-related macular degeneration, and glaucomatous optic nerve damage, in a single test.

Since its launch in global markets, the EyeArt system has been used by hundreds of clinics to enable safe and reliable eye screening for over 230,000 patients with diabetes and counting.

“I am particularly thrilled about this clearance as it showcases our extraordinary commitment towards our mission to screen every eye in the world to ensure timely diagnosis of life- and vision-threatening diseases,” said Kaushal Solanki, PhD, founder and CEO of Eyenuk. “The EyeArt system can now be used with multiple camera models in the U.S., which significantly expands access for patients who can be screened in their primary care doctor’s office for preventable blindness due to diabetes.”

According to the company, the latest FDA clearance for the EyeArt v2.2.0 system is based on clinical data from a prospective multi-center clinical trial (NCT04984200) that demonstrated its performance for the Topcon NW400 cameras:

• 94.4% sensitivity and 91.1% specificity for more-than-mild DR detection, and
• 96.8% sensitivity and 91.6% specificity for vision-threatening DR detection.

The clearance also paves the way for the EyeArt system to receive an upgraded image quality assessment module and Eyenuk’s proprietary Real-Time Image Quality Feedback solution.

Eyenuk’s Real-Time Quality Feedback solution improves the workflow of diabetic retinopathy screenings in primary care clinics by significantly limiting the need for patients’ eyes to be dilated. Dilation requires staff to administer eyedrops and then wait up to 15 minutes for the patient’s pupils to become enlarged before conducting a screening. Dilation is only necessary with the EyeArt AI system in the rare instance of the AI being unable to “grade” an image – that is, conclude that DR is either present or not present. With Real-Time Quality Feedback, the system operator is alerted to an inconclusive result before a report is produced and directed to capture a clearer image.

Reaction

According to Peter McDonnell, MD, director of The Wilmer Eye Institute, The Johns Hopkins University School of Medicine, Baltimore, and co-chief medical editor of Ophthalmology Times, the approval adds to the growing consensus that AI has an increasing role to play in the detection of eye disease in the future and adds to the previously approved devices serving this purpose.

Sharon Fekrat, MD, a professor of ophthalmology at Duke University School of Medicine, noted that ultra-widefield (UWF) imaging using the Optos California camera is another camera that does not require dilation, and obtaining steered images using an UWF camera can not only screen for diabetic retinopathy but other fundus pathology as well and may be the most fruitful.

Fekrat noted the FDA clearance for the Eyenuk has the potential to boost early detection of diabetic retinopathy.

“Widespread screening for diabetic retinopathy, irrespective of the camera or human vs AI read, would allow the earlier detection and treatment of diabetic retinopathy to perhaps obviate surgical intervention which is associated with more risk than nonsurgical interventions,” she said. “Earlier detection may increase the number of intravitreal injections and laser that is performed. Time will tell.”

Moreover, Fekrat noted the availability of automated diabetic retinopathy screening by primary care physician offices will likely have different effects in different parts of the country.

“In more rural areas, it would allow the detection of diabetic retinopathy in individuals who may not otherwise undergo screening due to a variety of factors. In more urban areas, such screening may have less utility since the number of eye doctors per capita is much greater,” she explained. “Automated widespread diabetic retinopathy screening may free up eye clinic spots for other patients with serious eye disease. Or it is possible that a patient with "any diabetic retinopathy" would then be referred to an ophthalmologist and this could fill up clinics with patients who otherwise not have sought out care.”

When it comes to the impact on workflow in a primary care or eye practice, Fekrat noted that for

newly diagnosed diabetics, primary care physicians can add retinal screening to their workup by obtaining same day fundus photographs in their office, if they have a camera available and someone who is available to obtain the images.

“Given the capital expense of obtaining a camera, however, it is a possibility that primary care physicians will not opt to perform retinal screening in their office but instead would rather send the patient to an eye doctor,” she explained. “With Medicare cuts, primary care physician billing receipts are decreasing and hence they are less likely to invest in a fundus camera and increase their time spent with the patient.

Fekrat noted that reimbursement for fundus photos is on the decline, and real world experiences will shine the light on the path forward in this space.

“Eye doctors can no longer bill for both extended ophthalmoscopy performed during a dilated eye exam AND fundus photographs obtained on the same day,” she concluded. “If more patients are detected to have diabetic retinopathy during primary care screenings and these patients are then sent to an eye doctor (usually a retina specialist) for further evaluation, the number of patients seeking care will increase. That can then pose challenges, especially when patient numbers in retina practices are increasing with the recent FDA approval of Syfovre.”