Explainable Artificial Intelligence for Diabetic Retinopathy and Diabetic Macular Edema Diagnosis: A Review

Retinal diseases are among the leading causes of visual impairment and blindness worldwide. Conditions such as diabetic retinopathy (DR), diabetic macular edema (DME), glaucoma, and age-related macular degeneration (AMD) significantly affect the quality of life of millions of individuals. Early diagnosis and timely treatment are crucial for preventing irreversible vision loss. However, manual examination of retinal images requires specialized ophthalmologists and is often time-consuming, especially in regions with limited healthcare resources. Recent advances in artificial intelligence (AI) and deep learning (DL) have demonstrated remarkable success in automated retinal disease diagnosis. Convolutional Neural Networks (CNNs), transfer learning approaches, and Vision Transformer (ViT)-based architectures have achieved high diagnostic accuracy using retinal fundus and optical coherence tomography (OCT) images. These methods can assist clinicians in screening large populations efficiently and consistently. Despite their impressive performance, most deep learning models operate as