Detection of Diabetic Retinopathy Using Convolutional Neural Network
8 Pages Posted: 15 Jul 2019 Last revised: 30 Sep 2019
Date Written: May 17, 2019
Diabetic Retinopathy(DR) is a disease which is caused due to long term diabetes. It is a visual exposition of diabetes, caused by impairment of the blood vessels in the retina. Around 80 percent of the population having diabetes for more than 10 or more years has some stage of the disease. In this paper, we propose a system that digitally detects the disease using Convolutional Neural Networks (CNN). CNN are used for the implementation of the proposed system as it most apposite for image data set. The programming language used for the implementation is python and libraries included are Keras, opencv, numpy, etc. The accuracy achieved by the model is 74.04 percent for 5 class classification. The accuracy can be further improved by increasing the size of the dataset as only a subset of the data set is considered for implementation of the model because of hardware constraints. The model is implemented on kaggle cloud.
Keywords: Diabetic Retinopathy, Convolutional Neural Network, Colour Fundus
JEL Classification: Y60
Suggested Citation: Suggested Citation