Glaucoma Prediction Using Deep Learning Techniques - A Comparative Study
P. Malin Bruntha, Iwin Thanakumar Joseph S, S. Saravanan, D. Narmadha, Suresh Subramanian, G. Naveen Sundar
Abstract
The degenerative eye disease glaucoma affects the optic nerve and is frequently brought on by elevated intraocular pressure. It results in permanent eyesight loss. Around 80 million individuals worldwide are impacted, with 10% of those developing blindness. To prevent visual impairment, early detection and care are essential. In this paper, deep learning architectures for the categorization of glaucoma and healthy eyes are analyzed. Various pre-trained deep-learning approaches are studied to categorize the fundus images of the retina into two categories: normal and glaucomatous. It is noted that the performance of EfficientNet is superior compared to other architectures in glaucoma prediction.