Validation and optimization of smart eye camera as teleophthalmology device for the reduction of preventable and treatable blindness in Nigeria.

Limited resources and staffing hinders efforts to reduce preventable blindness, especially in low- to middle-income countries. The slit-lamp examination (SLE), which is essential for ophthalmology practices, is often unavailable in primary and secondary eye care facilities due to the high costs and lengthy training required for operation. We conducted a cross-sectional, multicentre study exploring the potential for a smart eye camera (SEC; a tele-ophthalmology handheld device developed by OUI Inc., Japan) to address the limitations of the SLE. Ocular diagnoses, visual acuity assessments and examinations of the eyes were performed independently using both a conventional SLE and a SEC. Four independent assessors (blind to the study) reviewed the images captured by the SEC and the SLE as administered by separate investigators. All analyses were performed using R version 4.2.2 for macOS at a 5% level of statistical significance. The results of the image quality analysis demonstrated that the number of higher-quality images was significantly higher (p < 0.05) for the images captured using the SEC device compared to the SLE machine. Remarkably, up to 96% accuracy of diagnosis was recorded with SEC. Evaluation of diagnostic accuracy rates derived from images obtained from both machines revealed a degree of divergence in assessments among evaluators, yielding a Fleiss's Kappa value of 0.092. The sensitivity analysis for the SEC device revealed a reasonably strong capacity to correctly identify true positive cases, with an average sensitivity score of 90%. The results of this study indicate that SEC can effectively evaluate anterior segment lesions in ophthalmology. © 2024. The Author(s), under exclusive licence to The Royal College of Ophthalmologists.