Handheld Fundus Camera for Diabetic Retinopathy Screening: A Comparison Study with Table-Top Fundus Camera in Real-Life Setting.

The aim of the study was to validate the performance of the Optomed Aurora handheld fundus camera in diabetic retinopathy (DR) screening. Patients who were affected by diabetes mellitus and referred to the local DR screening service underwent fundus photography using a standard table-top fundus camera and the Optomed Aurora handheld fundus camera. All photos were taken by a single, previously unexperienced operator. Among 423 enrolled eyes, we found a prevalence of 3.55% and 3.31% referable cases with the Aurora and with the standard table-top fundus camera, respectively. The Aurora obtained a sensitivity of 96.9% and a specificity of 94.8% in recognizing the presence of any degree of DR, a sensitivity of 100% and a specificity of 99.8% for any degree of diabetic maculopathy (DM) and a sensitivity of 100% and specificity of 99.8% for referable cases. The overall concordance coefficient k (95% CI) was 0.889 (0.828-0.949) and 0.831 (0.658-1.004) with linear weighting for DR and DM, respectively. The presence of hypertensive retinopathy (HR) was recognized by the Aurora with a sensitivity and specificity of 100%. The Optomed Aurora handheld fundus camera proved to be effective in recognizing referable cases in a real-life DR screening setting. It showed comparable results to a standard table-top fundus camera in DR, DM and HR detection and grading. The Aurora can be integrated into telemedicine solutions and artificial intelligence services which, in addition to its portability and ease of use, make it particularly suitable for DR screening.