Journal Pre-proof

SBFI uses smartphone cameras, usually combined with optical adapters and illumination systems, to capture fundus images. SBFI systems can rely on high- resolution sensors in smartphones, however, older models with inferior image quality and lower image resolution were also suitable for fundus imaging. Most SBFI approaches rely on external magnification lenses in addition to the smartphone’s own optical system, simulating the function of a traditional indirect ophthalmoscope. Yet, there are a few technical approaches for SBFI where the concept of direct ophthalmoscopy is utilized without the need of an additional lens. A built-in or external light source is needed for proper retinal illumination. Utilization of the smartphone’s light source can complicate medico-legal certification, as, by this, the smartphone becomes a medical device emitting light on the retina. Some systems use 3D-printed adapters, clip-on lenses, or beam splitters to align the smartphone camera, the illumination source, and the external lens.

This review was conducted as a scoping review with a structured literature search. The primary bibliographic search was performed in the web of science core collection Journal Pre-proof (search details in Supplement Appendix A.1), last updated on 6th of December 2025, screened 1979 publications for relevance, removed 51 duplicates and identified 363 potentially relevant articles through title screening, out of which after abstract screening and detailed revision 193 were included in our Review. 81 relevant articles and publications have additionally been included by cross-referencing and grey literature research.

The combination with artificial intelligence (AI) has also produced initial promising results, which might further increase efficiency and potential cost-effectiveness. Nevertheless, challenges ranging from image quality and limited fields of view to regulatory, data privacy, and medico-legal hurdles remain, emphasizing the importance of further research and structured implementation of SBFI programmes

Over the past decade, SBFI technologies have steadily improved and have been evaluated for screening, referral triage and teaching applications. However, their current use remains limited relative to the global need for retinal imaging, and evidence of sustained large-scale impact on eye-care delivery remains scarce. Our review indicates that SBFI may support fundus imaging workflows, particularly in resource-limited settings where access to conventional imaging equipment and specialist care is limited. Its main current value lies in expanding access to image acquisition and telemedical referral pathways for conditions such as diabetic retinopathy, glaucoma suspect identification and referral, and selected retinopathy of prematurity applications. The integration of AI may further support scalability through image-quality assessment, workflow support, and automated grading, but its impact in clinical practice will depend on robust validation, implementation, and access to downstream care.