Classifying Normal Eye, Cataract Eye, Glaucomic Eye and Diabetic Retinopathic Eye using ResNet50 ML algorithm

The diagnosis and classification of ophthalmic conditions are critical for effective treatment and management. While traditional diagnostic methods rely on expert interpretation of clinical signs and imaging, there is a growing need for automated, accurate, and rapid systems. This paper targeted to exploit the prowess of machine learning (ML) and deep learning (DL) algorithms in classifying four significant eye conditions: normal eyes, cataract, glaucoma, and diabetic retinopathy. We trained ResNet50 model in sequential mode to tell the difference between these conditions by using large datasets that included clinical features and retinal images. Our results indicate that ML and DL algorithms can achieve a high degree of accuracy and could serve as valuable tools for screening and diagnostic purposes and possibly could save a huge amount of time and lives.