Diabetic Retinopathy Diagnosis with Ensemble Deep-Learning

Abstract

Diabetic retinopathy is an eye disease, is a condition in which retina is damaged due to diabetes mellitus. It is a major cause of blindness. Although many artificial intelligence methods has been applied to diabetic retinopathy diagnosis. This method is a new approach in this problem domain. It's early detection can help in tackling the future damages due to the disease. Here, model is ensemble of pre-existing GoogleNet, AlexNet and ResNet50 architectures. The best machine learning models that were used was GoogLeNet achieving highest accuracy for this job. Here, The results are standing out with the GoogLeNet's accuracy.