Diabetic Retinopathy Prediction Using CNN Models

Abstract

Diabetic Retinopathy (DR) is an eye condition that mainly affects individuals who have diabetes and is one of the 
important causes of blindness in adults. As the infection progresses, it may lead to permanent loss of vision. 
Diagnosing diabetic retinopathy manually with the help of an ophthalmologist has been a tedious and a very 
laborious procedure. This paper not only focuses on diabetic retinopathy detection but also on the analysis of 
different DR stages, which is performed with the help of Deep Learning (DL) and transfer learning algorithms. 
ResNet 50, EfficientNet B0, and DenseNet 121 are used on a huge dataset with around 3662 train images to 
automatically detect which stage DR has progressed. Five DR stages, which are 0 (No DR), 1 (Mild DR), 2 
(Moderate), 3 (Severe) and 4 (Proliferative DR) are processed in the proposed work. The patient’s eye images are fed 
as input to the model. The proposed deep learning architectures like ResNet 50, EfficientNet B0, DenseNet 121, are 
used to extract the features of the eye for effective classification. The models achieved an accuracy of 97.34%, 94.88% 
and 80.77% respectively. The paper concludes with a comparative study of the ResNet 50, EfficientNet B0, DenseNet 
121 architectures that highlights ResNet 50 as the perfect deep learning classification model for automated DR 
detection. Make the above abstract more professional.