Classification of Diabetic Retinopathy using Pre-Trained Deep Learning Models

TL;DR

The paper tackles automated five-class diabetic retinopathy classification from fundus images. It evaluates a CNN baseline against fine-tuned pre-trained networks (VGG16, MobileNet, InceptionV3, InceptionResNetV2) on a Kaggle DR dataset with two input resolutions, using data augmentation and a Flask-based deployment to demonstrate practical accessibility. MobileNet achieves the highest AUC around 0.70, with InceptionResNetV2 close at 0.69, illustrating the advantage of transfer learning for medical imaging tasks. The work highlights a path toward mobile-ready DR assessment and provides insights into model selection for imbalanced, high-detail medical imagery.

Abstract

Diabetic Retinopathy (DR) stands as the leading cause of blindness globally, particularly affecting individuals between the ages of 20 and 70. This paper presents a Computer-Aided Diagnosis (CAD) system designed for the automatic classification of retinal images into five distinct classes: Normal, Mild, Moderate, Severe, and Proliferative Diabetic Retinopathy (PDR). The proposed system leverages Convolutional Neural Networks (CNNs) employing pre-trained deep learning models. Through the application of fine-tuning techniques, our model is trained on fundus images of diabetic retinopathy with resolutions of 350x350x3 and 224x224x3. Experimental results obtained on the Kaggle platform, utilizing resources comprising 4 CPUs, 17 GB RAM, and 1 GB Disk, demonstrate the efficacy of our approach. The achieved Area Under the Curve (AUC) values for CNN, MobileNet, VGG-16, InceptionV3, and InceptionResNetV2 models are 0.50, 0.70, 0.53, 0.63, and 0.69, respectively.

Figures (1)

• Figure 1: Diabetic Retinopathy Stages