Deep Learning-Based Multi-Class Classification of Retinal Diseases from OCT Images Using Transfer Learning

Optical Coherence Tomography (OCT) is a non-invasive imaging modality widely employed for retinal disease diagnosis. However, manual interpretation of OCT images is time-consuming, subjective, and requires expert ophthalmological knowledge. This paper presents a deep learning-based framework for the automated classification of retinal diseases, including Diabetic Macular Edema (DME), Choroidal Neovascularization (CNV), Drusen, and Normal retina. The proposed system integrates image preprocessing, transfer learning-based feature extraction, and multi-class classification using pretrained convolutional neural networks. Three different model architectures—VGG-16, ResNet-50, and a hybrid CNN-feature fusion model—are evaluated on the Kermany OCT dataset. Experimental results demonstrate that the ResNet-50 model achieves the highest classification accuracy of 97.83%, outperforming VGG-16 (96.12) and the hybrid model (95.04%). Compared to previous studies reporting accuracies between 94.5% and 97.5%, the proposed approach shows improved diagnostic performance. The results highlight the potential of deep learning-based OCT analysis to assist ophthalmologists in early disease detection and clinical decision-making.