DIAGNOSIS OF DIABETIC RETINOPATHY USING CONVOLUTION NEURAL NETWORKS
Diabetic Retinopathy is a complication of diabetes, especially for those with type 2 diabetes. High blood sugar levels over a period of time can damage the blood vessels in the retina, making them, swell and leak. In some cases it may also happen that they block blood from passing through. These lead to development of irregular fresh blood vessels in the retina. All of these conditions affect the vision adversely leading ultimately to loss of vision. The early stages of diabetic retinopathy are painless and symptom less, and hence can go undetected for a long time. It is therefore recommended for diabetics to have an annual eye fundus examination. As the disease advances certain symptoms may occur like: Sudden changes in vision, Reduction in night vision, Distorted vision, Impaired colour vision, Eye pain.Identification of Diabetic retinopathy at early stage is useful for for clinical treatment. Researchers already proposed several feature extraction techniques to classify the retinal images with and without diabetic retinopathy but still classification technique is still complex task for retinal image. Deep learning is widely used in numerous applications one of them being medical analysis. Feature extraction and Image classification are considered to be the most popularly used approaches done using deep learning process. In this proposed work, we used deep learning technique namely convolutional Neural Network (CNN) an efficient model for detecting Diabetic Retinopathy by preprocessing digital fundus images and further segmenting it for feature extraction. The feature extraction of the images is done by training the convolution neural networks to classify whether the image is affected or not affected by Diabetic retinopathy. The effectiveness of the proposed model is assessed which produces 81.27% affectability, 99.91% explicitness, and 99.71% precision. The fulfillment of the model is progressively precise when contrast with existing as it makes use of CNNs to train and validate the data set.