Non-proliferative Diabetic Retinopathy(NPDR) using Deep learning approach

Non-proliferative diabetic retinopathy (NPDR), a frequent eye problem in diabetics, can result in vision loss. For prompt treatment and prevention, early detection is essential. Promising outcomes in NPDR detection have been observed in recent breakthroughs in computer vision and deep learning systems. In order to detect NPDR utilizing retinal fundus pictures, a unique deep learning-based approach is proposed in this study. The approach employs a fully connected layer for classification and a pre-trained convolutional-neural-network for extraction of feature. The algorithm performs better than state-of-the-art techniques by accuracy, specificity, sensitivity, and under the (ROC) receiver operating characteristic curve, according to experiments conducted on a publicly accessible dataset.