Evolution and Testimony of Deep Learning Algorithm for Diabetic Retinopathy Detection

Abstract

A deep neural network (DNN) is an artificial neural network (ANN) with various layers incorporated between the input and output layers. Deep Neural Networks (DNNs) embodies such a type of network where each and every respective layer performs convoluted functions such as demonstration and conceptualization that comprehend images, sound and text. Deep learning functions along with artificial neural networks, these networks are well suited for imitating how humans contemplate and learn. DL mainly consists of analyzing, learning and improving on its own by inspecting computer algorithms. Image classification, language translation, and speech recognition has been supported by deep neural networks. Without human involvement, deep neural networks can resolve any pattern recognition problem. This paper, focuses on the growth and confirmation of a deep Learning algorithm for early-stage diabetic retinopathy detection. Diabetic Retinopathy, a disorder that crops up in the human eye, if not treated at an early stage, may lead to blindness by lesions on the retina. Hence, ResNet 50 a deep learning technique is presented to automate the recognition of the diabetic retinopathy images by the classification of retinal fundus images from kaggle database. Over 3662 images which are retinal fundus images are used for training and validation. The accuracy achieved after first epoch is 90.74% and after final epoch 91.60%.