Alzu'bi et al. investigated the use of deep learning models to automatically classify optical coherence tomography (OCT) images for distinguishing between diabetic macular edema (DME), age-related macular degeneration (AMD), and normal retinal conditions. The research was conducted using a dataset of 1,040 OCT images from King Abdullah University Hospital, with images preprocessed, augmented, and segmented to enhance model performance. Three pretrained convolutional neural networks (ResNet152, InceptionV3, and MobileNetV2) were evaluated, and model interpretability was assessed using Grad-CAM visualization techniques. The benchmarking results on the OCT dataset showed that InceptionV3 achieved the highest accuracy at 97%, with perfect classification of AMD cases (F1 = 1.00) and strong performance across all classes (F1 ≥ 0.95), making it the most reliable model. ResNet152 followed closely with 95% accuracy, showing balanced metrics (F1 = 0.99 for AMD, 0.93 for DME and Normal). Alzu'bi et al. highlight that a developed deep learning framework, using InceptionV3 and ResNet152 combined with AI techniques, can effectively address the diagnostic challenge of accurately differentiating between DME, AMD, and normal retinal conditions.