Automated Deep Learning Framework for Retinal Disease Classification from OCT Images: A Large-Scale Study Across Four Categories

OCT imaging, which produces high-resolution cross-sectional images of the retina that are vital for identifying a variety of retinal illnesses, has emerged as a critical diagnostic technique in ophthalmology in recent years. However, it takes a lot of effort and specialized knowledge to analyze OCT pictures manually, hence automated techniques are needed to help doctors. This work introduces a deep learning-based framework that uses a sizable dataset of 84,495 retinal OCT pictures to automatically classify the images into four different illness categories. Our model provides a highly accurate, effective approach for automated diagnosis by utilizing convolutional neural network (CNN) architectures to efficiently capture the complex characteristics present in OCT pictures. High classification accuracy is attained by the suggested approach, indicating its resilience and ability to expedite the diagnostic procedure.