Diabetic Retinopathy Diabetic retinopathy (DR) is a medical condition in which damage occurs to the retina due to diabetes mellitus. It is a leading cause of blindness. Diabetic retinopathy affects up to 80 percent of those who have had diabetes for 20 years or more. Diabetic retinopathy often has no early warning signs. Retinal screening aids in the early detection and treatment of DR. The need for a comprehensive and automated method of DR screening has long been recognized, and previous efforts have made good progress using image classification, pattern recognition, and machine learning. The objective of this project is to develop High End-end Deep learning framework technology to improve Diabetic retinopathy screening. In this project, we use Convolutional Neural Networks (CNN), Transfer Learning and Multi-task Learning Techniques to provide a deep-learning-based solution for stage identification of diabetic retinopathy by single photography of the human fundus and ideally improving performance resulting in models with realistic clinical potential.
Problem statement To build a machine learning model to speed up disease detection and to help classify diabetic retinopathy into 5 different stages.
Kaggle Dataset: https://www.kaggle.com/c/aptos2019-blindness-detection
Data Description : Large set of retina images taken using fundus photography under a variety of imaging conditions. A clinician has rated each image for the severity of diabetic retinopathy on a scale of 0 to 4:
0 - No DR 1 - Mild 2 - Moderate 3 - Severe 4 - Proliferative DR
It comprises of 3662 photos of eyes tagged with the five severity stages of diabetic retinopathy, which might be used to train a machine learning model to solve the problem. There are 3662 retina images in the dataset, all of which of various sizes. Only the ground facts of the training photos are provided to the general audience. Furthermore, 1805 of the photos are normal, whereas 1857 are DR.
