A comprehensive review of portable retinal cameras: Technical features, ai integration, and clinical potential

Purpose: To evaluate and compare commercially available portable retinal cameras with a focus on technical specifications, clinical applications, and the integration of artificial intelligence (AI) for ophthalmic screening, especially in low- and middle-income countries (LMICs). Design: Systematic review of the literature. Methods: Systematic searches of PubMed and OpenAlex were conducted up to September 2025, without language restrictions, using terms such as portable retinal camera, handheld retinal camera, and smartphone-based fundus camera. Devices were included if they were commercially available and described in peer-reviewed publications with technical or clinical data. Prototypes and systems relying solely on external smartphone lenses without integrated optics were excluded. Data extracted included imaging specifications, ergonomics, power sources, AI functionalities, quality control features, and reported clinical applications. Devices were categorized as smartphone-attached or standalone handheld systems. Results: The search retrieved 870 records (PubMed = 277; OpenAlex = 593). After removing duplicates and screening, 509 articles were included in the review, collectively reporting on 38 portable retinal cameras, of which 17 were commercially available. The most frequently reported devices were the Volk Pictor Plus, Phelcom Eyer, Optomed Aurora, ZEISS VISUSCOUT 100, and Remidio FOP. Smartphone attached systems offered greater portability and affordability, whereas standalone handheld systems provided integrated functionality, higher imaging stability, and smoother clinical integration. AI features varied across devices, encompassing referable diabetic retinopathy detection, abnormality triage, systemic risk prediction, and automated image-quality assessment. Clinical applications extended beyond diabetic retinopathy and retinopathy of prematurity to include glaucoma, AMD, and exploratory use in systemic conditions such as sepsis and COVID-19. Conclusion: Portable retinal cameras already demonstrate clear utility in extending ophthalmic screening and diagnostic services, particularly for diabetic retinopathy and retinopathy of prematurity, while also showing potential in broader clinical and systemic applications. Their portability and cost-effectiveness make them valuable for outreach and telemedicine programs, especially in LMICs. The integration of artificial intelligence further enhances their functionality, though variability in device design, regional availability, and regulatory status highlights the need for standardized validation, recurring local assessments, and head-to-head comparative studies. Real-world evaluations remain essential to ensure effective, safe, and equitable deployment.