Technical Feasibility Study for a Retinal Camera System

Background/ObjectivesImprovements in diabetic retinopathy (DR) screening could boost early detection rates and contribute to better patient outcomes. The availability of primary care-based DR screening could address this issue, although operational challenges remain. This study represents a preliminary evaluation of the feasibility and operational performance of a retinal screening system suitable for integration into a primary care setting. MethodsThis was a multi-center, data collection study for a non-mydriatic retinal camera system, conducted in 4 iterative development phases; phase 4 simulated a clinical workflow. Study endpoints evaluated image quality (gradability) and operational characterization in phase 4 (time to complete image capture). Participants were required to be at least 22 years of age, and separate participant pools were recruited throughout study phases. In each phase, multiple images per eye were collected, and graded independently by sponsor-employed and external graders. ResultsThe study included 212 participants. In phases 1-3 (n=192), evaluation of the first image captured per eye for all participants (k=383) by non-sponsor graders showed 93.7% of images (359/383) as gradable; in phase 4 (n=20; k=55), there were 81.8% (45/55) gradable images. Regarding operational results, 90.1% of first-attempts trigger times were less than 30 seconds; by the third attempt, 96.7% were less than 30 seconds, ConclusionsThese are encouraging preliminary results in terms of performance and usability for the retinal camera system in this study, which may have potential application in primary care clinics.