Detection of Retinal Disease Using Deep Learning Architecture

World Health Organization (WHO) estimates indicate around 2.2 billion individuals across the globe face challenges related to their eyesight. Vision impairment may have been avoided in at least 1 billion of these cases, or it has not yet been treated. Disease-related biomedical image classification is a laborious, manual process. On the other hand, machine intelligence is now a more practical approach for the detection and classification of medical images thanks to recent advancements in computer vision. This study looks into the application of deep learning models like as CNN, Xception, and VGG19 for the early diagnosis of eye illnesses using medical images. The 83,484 retinal OCT pictures of patients with different tunnel dis-ease-related conditions are used to train these models. The CNN model has the highest accuracy, with 95.47%, 85.69%, and 87.15%, respectively. Improving early detection and treatment is the goal to lower healthcare costs and better patient outcomes.