Detection of diabetic Retinopathy using Retinal Fundus Images

Abstract

Diabetic retinopathy is one of the diabetes consequences that affects the eyes. This is caused by damage to the blood vessels in the retina, the light-sensitive tissue in the rear of the eye. It may create no symptoms at first, or it may cause minor eyesight difficulties. When the blood vessels become damaged, they may leak and this leakage can cause dark spots on our vision. The DR can be detected by finding the Hard Exudate present in it. The deep networks are becoming deeper and more complex. So that adding more number of layers to a neural network can make it stronger for image related tasks. But the main disadvantage in adding more layers is that, it may greatly reduces the accuracy of the image and also the data models are complex. In order to overcome this drawback, Recurrent Neural Network can be introduced. The fundamental benefit of using a recurrent neural network is that it can represent a collection of data in such a way that each pattern may be presumed to be reliant on the one before it. It can process inputs of any length. Even if the input size is large, the model size will not change. It makes the training process faster and attains more accuracy while compared to other neural networks. This method greatly reduces the loss of accuracy because each layer knows the information of the top layers while updating the weights. This Recurrent Neural Network has more number of parameters , so it is obvious that it can produce better result as compared to other net.