Implementation of Complete Glaucoma Diagnostic System Using Machine Learning and Retinal Fundus Image Processing

Le résumé fourni par la source

Glaucoma is one of the leading diseases causing irreversible blindness worldwide. Nearly 70 million individuals suffer from glaucoma globally in 2020 and by 2040 this number is expected to rise to 111.8 million. Moreover, half of the patients affected by glaucoma remain undiagnosed until a relatively late stage due to the slow and asymptomatic nature of the disease in its earlier stages. Early detection of glaucoma based on quality images is highly needed. Examining retinal fundus image is one of the popular screening approach for glaucoma. However, this approach is laborious with specific technological devices and requires eye specialists. It is therefore difficult to conduct in large scale of a country, especially with provincial medicine centers. In this paper, we introduce a complete real-time system for glaucoma classification, which can support eye doctors to diagnose the disease potential efficiently. Applying latest technologies, we build a complete system including IoT device for retinal funus image processing, machine learning on cloud for glaucoma classification, and an desktop application for glaucoma diagnostic. According to our review, this is the first complete system in this field with the classification accuracy of 85%.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.