Diabetic Retinopathy Detection using Pretrained Models of Deep Learning

Diabetic Retinopathy (DR) is a type of disease related to eye caused by diabetes. It is one of the reason for blindness. Detecting retionopathy is very essential process in reducing partial blindness of patients. Identifying DR consumes time and involves lot of manual process by opthamologist. Detection of DR helps in taking quick decisions. This paper discloses automated binary classification system over 388 retina images(healthy and severeDR).This automated classification system is based on pretrained models VGG-16 and VGG-19 of deep learning with convolution neural network.Though there are many intelligent system ,clarity in the images plays a vital role.This automated classification system is build with minimal code using pretrained model and KNIME analytical platform. The proposed system achieves around 91% with both VGG16 and VGG-19 with fewer image preprocessing steps