The worldwide burden of visual impairment disproportionately affects people residing in low and middle-income countries (L&MICs). In addition, many of the causes of visual impairment are preventable. The Vision 2020 Right for Sight initiative set the challenge to eliminate avoidable blindness by 2020. One of the key elements of the Vision 2020 initiative is infrastructure strengthening, which includes technology and software development for telemedicine eye care. In response several novel low cost retinal imaging devices have been developed along with software to montage and analyse retinal images.

Many causes of blindness can be diagnosed early and prevented if appropriate retinal imaging technology is available in these low resource settings. Typically these diagnostic devices are unavailable to health care workers in L&MICs due to high initial cost and difficulty in maintaining the devices. Even if a health care worker has access to devices they may not be able to interpret the clinical signs appropriately and make onward referral for treatment. As a result many individuals with diseases such as glaucoma, diabetic retinopathy and retinopathy of prematurity present with advanced and untreatable disease and never access care.

This study aims to analyse the structure of optic nerves from previously acquired videos of healthy medical students using three of these new novel retinal imaging devices; Arclight, D-eye and Epicam C. 

This study aims to detect eye diseases by analyzing the retina images with a mobile app. This app allows retina images acquisition and optic nerve hear structure on a mobile phone platform using  new novel retinal imaging devices; Arclight, D-eye and Epicam C. These small devices are cheaper, lighter and easier and can be mounted easily on top of any smart phone for capturing images by any individual. 

By applying image processing and machine learning methods, a complete shape of retina can be reconstructed for different eye disease detection. For example, Glaucoma is a disease of the optic nerve that can lead to blindness. Early diagnosis by identifying structural changes in the optic nerve is key to preventing needless disability. By designing an automatic image segmentation model, which can identify these characteristic structural changes in the optic nerve, assists in early diagnosis and monitoring of the disease. 

This study has been started last year at the University of St Andrews in collaboration with David Harris-Birtill, Dr Andrew Blaikie, Joseph Upaa and Mr Teng Zuo. The project had good outcome so far but still an ongoing project and we hope to make the complete app available for helping people who have not enough access to clinicians in undeveloped countries. Also such an application will help clinicians with limited access to expensive facilities better understand potentially blinding diseases and access opinions from remote experts.