A mathematical model based on machine learning and deep learning forcategorizing ocular diseases

Abstract

 The prevalence of ocular disorders is a global issue that affects everyone. If left untreated or undetected, several eye disorders, including cataracts, glaucoma, and diabetic retinopathy, can lead to blindness or impaired vision. Numerous studies have revealed that there is a general lack of knowledge about ocular disorders and how seriously they can affect a person\textquotesingle s quality of life. Governmental agencies and a few NGOs have stepped up throughout the years to raise awareness of the need to maintain a regular eye care regimen and attend checkups among the general public. However, identifying these disorders takes time and frequently requires manual evaluation by highly qualified ophthalmologists. The goal of this study is to develop a model that can help ophthalmologists quickly detect serious eye conditions such as cataracts, glaucoma, and diabetic retinopathy. The most accurate model among three different machine learning and deep learning models is used to categorise various eye conditions. Metrics including the overall accuracy score, precision, recall score, and f1 score are used to evaluate performance. Notably, 52\%, 59\%, and 61\%, respectively, are the Accuracy ratings for the Random Forest classifier, Convolutional Neural Network, and XGBoost.