Diabetic Retinopathy Detection using Deep Learning
Supriya Mishra, Seema Hanchate, Zia Saquib
Abstract
Diabetic Retinopathy (DR) is human eye illness which occurs in individuals who have diabetics which harms their retina and in the long run, may lead visual deficiency. Till now DR is being screened manually by ophthalmologist which is a very time consuming procedure. And henceforth this task (project) focuses on analysis of different DR stages, which is done with Deep Learning (DL) and it is a subset of Artificial Intelligence (AI). We trained a model called DenseNet on an enormous dataset including around 3662 train images to automatically detect the DR stage and these are classified into high resolution fundus images. The Dataset which are using is available on Kaggle (APTOS). There are five DR stages, which are 0, 1, 2, 3, and 4. In this paper patient's fundus eye images are used as the input parameters. A trained model (DenseNet Architecture) will further extract the feature of fundus images of eye and after that activation function gives the output. This architecture gave an accuracy of 0.9611 (quadratic weighted kappa score of 0.8981) to DR detection. And in the end, we are comparing the two CNN architectures, which are VGG16 architecture and DenseNet121 architecture.