AI-Powered Portable Ophthalmic Fundus Imaging Device for Community Eye Disease Screening

Fundus examination is the core diagnostic method for ocular fundus diseases such as diabetic retinopathy, glaucoma, macular degeneration and hypertensive retinopathy. Traditional desktop fundus cameras are large in size, expensive in price and require professional ophthalmologists to operate and interpret images, which are only equipped in large-scale tertiary hospitals. Community health service centers and grassroots medical institutions lack professional fundus screening equipment and technical personnel, resulting in low coverage of grassroots eye disease screening and serious missed diagnosis of early fundus lesions. To solve the problem of uneven distribution of ophthalmic medical resources, this study develops an AI-powered portable fundus imaging diagnostic medical device suitable for grassroots community screening. The device is lightweight and portable, supports non-mydriatic fundus photography, and can complete high-definition fundus image acquisition in 2 seconds without pupil dilation, avoiding the discomfort caused by mydriasis drugs. Built-in optimized deep learning diagnostic algorithms can automatically identify and grade common fundus lesions, including microaneurysms, hemorrhage, exudation and macular edema, and generate standardized screening reports in real time. The algorithm model is trained based on millions of clinical fundus images, with a diagnostic accuracy of 96.5% for common fundus diseases. Community screening practice shows that the portable intelligent fundus device is simple to operate, suitable for non-professional medical staff to use, and can complete large-scale rapid eye disease screening. It fills the gap of grassroots fundus screening equipment, improves the early diagnosis rate of chronic eye diseases, and promotes the hierarchical diagnosis and treatment of ophthalmic diseases.