A Survey on Artificial Intelligence and Deep Learning Techniques for Diabetic Retinopathy Detection and Classification
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Diabetic Retinopathy (DR) is a leading cause of preventable blindness, and early diagnosis remains challenging due to subtle lesion patterns and the need for expert-grade annotations.Recent advances in Artificial Intelligenceparticularly Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs)-have significantly improved automated DR screening accuracy.This survey provides a structured analysis of existing machine learning and deep learning approaches, highlights commonly used datasets, preprocessing strategies, and evaluation metrics, and synthesizes comparative results across major studies.In addition to summarizing current progress, the survey identifies limitations including dataset imbalance, poor model interpretability, and privacy concerns in clinical deployment.Future directions emphasize ViT-XAI integration, multi-modal fundus-OCT frameworks, federated ViT architectures, and privacy-preserving learning to enable reliable, scalable, real-time DR diagnosis.
Fields of Study & Subjects
Retinal Imaging and Analysis
Retinal Diseases and Treatments
Advanced Neural Network Applications
Publication details
- Journal / source
- Advances in intelligent systems research/Advances in Intelligent Systems Research
- Pages
- 212–222
- Publication date
- 1 April 2026
- Language
- EN
- Document type
- Book chapter