Detection of Diabetic Retinopathy Using Deep Learning

• Version
• Download 4
• File Size 0.00 KB
• File Count 1
• Create Date 14 March, 2026
• Last Updated 14 March, 2026

Detection of Diabetic Retinopathy Using Deep Learning

Shubham Vartak, Ajinkya Parte, Soham Ajgaonkar, Hitesh Ghagave

DOI : 10.46335/IJIES.2023.8.4.1

Abstract— Diabetic retinopathy is a major problem worldwide and many people are losing their vision because of it. The disease gets severe if it is not treated properly at its early stages. In this disease, the retinal blood vessel gets damaged due to high blood sugar levels which eventually blocks the light that passes through the optical nerves, making the patient with Diabetic Retinopathy blind. Diabetic Retinopathy is detected using manual screening, but this requires a skilled ophthalmologist which may not be available everywhere and thus diagnosis takes a lot of time. Therefore, we decided to build a deep learning model using which we will be able to detect multiple stages of severity for Diabetic Retinopathy. So, we studied and built widely-discussed models - Support Vector Machine (SVM) and Convolutional Neural Network (CNN) - and conducted a comparative study to determine the most suitable model. We found that CNN outperformed SVM in terms of accuracy and efficiency, making it the most suitable model for detecting multiple stages of severity for Diabetic Retinopathy. Thus, this automatic diabetic retinopathy detection model built using CNN can replace manual screening, enabling ophthalmologists to focus on patient care. Additionally, this model can assist inexperienced ophthalmologists in accurately diagnosing diabetic retinopathy.