Efficient diabetic retinopathy diagnosis through U-Net – KNN integration in retinal fundus images

Diabetic retinopathy (DR) is a retinal disorder that may lead to blindness in people all over the world. The major cause of DR diabetes for longer period and early detection only solution prevent vision. This paper focuses on classes Normal eye (No DR), Mild NPDR (Non-Proliferative Retinopathy), Moderate NPDR, Severe PDR. On fundus images, an effective method identifying diabetic proposed by combining U-Net architecture with K-nearest neighbours (KNN) algorithm. used segmenting exudates pictures, KNN algorithm final classification. combination enables accurate feature extraction efficient classification, effectively overcoming computational challenges common deep learning models. experiments are carried out utilizing publicly available dataset images from Kaggle assess effectiveness our suggested strategy. provides precise output when compared other models GoogleNet, ResNet18, VGG16. model training accuracy 82.96% PDR high short which prevents loss vision stage.