Diabetic Retinopathy Screening Telemedicine System Using Hybrid Function

Abstract

Diabetic retinopathy is a prevalent disease affecting the retinas of diabetic patients, often leading to blindness in adults. The condition progresses through various stages, with early detection posing a challenge. To address this issue, an imaging system using hybrid machine learning to detect diabetic retinopathy early while reducing the cost of this procedure and taking into account high efficiency. The system is based on imaging using a smartphone, lifting the lens with one hand and bringing it closer to the eye, and lifting the phone with the other hand while playing the video a lens attachment and a phone flash for illumination, followed by image enhancement and classification using hybrid machine learning. A telemedicine website was implemented to share the results from the main hospital for diagnosis to the sub-center remotely.