|Soon, an app may be able to help providers determine whether a patient has AMD. Photo: Retina; Wikimedia Commons; iPhone: Marianne Krohn on Unsplash. Click image to enlarge.|
In the near future, patients may be able to screen for and monitor AMD with their smartphones. Experts say real-time, self-administered screening could boost early detection among the elderly, whose limited mobility may contribute to delays in disease diagnosis.
Researchers recently proposed an automated, smartphone-based screening method to aid in early detection of AMD, and their study demonstrated that the approach is portable, cost-effective and less time-consuming.
The method for detecting drusen involves a smartphone and a small portable fundus camera. Volk-N-View and Welch-Allyn Panoptic Ophthalmoscope lenses were used in the study. The lenses snap onto a smartphone and provide a field of view of the fundus between 45° and 50°. First, a retinal image is analyzed to locate the macula, and the image is enhanced. Its intensity representation is improved, and then the software extracts features of interest and classifies the image as AMD-affected or healthy. Results achieved an average accuracy of 100%, 95.2%, and 94.3% against three different databases.
“Regular eye screening helps diagnose AMD and may prevent vision loss in the elderly,” the researchers wrote in their paper, published in a bioengineering journal. “Our proposed method seems like a good solution for real-time AMD screening on mobile devices.”
Practical application is still a ways away, and the involvement of diagnostic lenses essentially will require use in a medical setting. But the authors consider their work proof of concept and a foundation upon which to develop possible clinical implementations.
This work could provide a “computer-aided diagnosis” system for eye doctors “to take advantage of its mobility, cost-effectivity, detection performance and reduced execution time,” they wrote, noting that a commercialized option could be used to decrease the eyecare burden in underserved areas both globally and in rural areas of the United States.
Ben Sayadia S, Elloumi Y, Kachouri R, et al. Automated method for real time AMD screening of fundus images dedicated for mobile devices. Med Biol Eng Comput. March 18, 2022. [Epub ahead of print].