Prediction of Diabetic Retinopathy using Convolutional Neural Networks

Diabetic retinopathy (DR) is among the most dan-gerous diabetic complications that can lead to lifelong blindness if left untreated. One of essential difficulties in DR early discovery, which crucial for therapy progress. The accurate diagnosis stage famously complicated and demands a skilled analysis by expert being fundus images. This paper detects classifies its using retina images applying conventional neural networks transfer learning models. Three deep models were investigated: trained from scratch CNN pre-trained InceptionV3 Efficient-NetsB5. Experiment results show proposed model outperformed with 9 25% relative improvement F1-score compared EfficientNetsB5, respectively.