Eye Disease Classification and Detection using Deep Learning

Diabetic retinopathy, glaucoma, and cataract are the most common eye diseases that are difficult to detect manually in their early stages. If these diseases are not detected early, they can cause permanent vision loss. Manual detection of eye diseases may result in incorrect diagnoses. This project aims to build an efficient model for eye disease classification and detection using deep learning techniques. Machine learning algorithms were previously used for eye disease classification and detection. Support vector machines, convolutional neural networks, and ResNet50 were previously used for eye disease detection. These existing models are complex and work efficiently only for small datasets. The primary objective of this project is to predict whether a person has eye disease or not based on his or her retinal fundus imaging data. The proposed system will use the Visual Geometry Group 19 deep learning algorithm for eye disease classification and detection. The proposed system is to enhance the test accuracy and work with larger datasets. The proposed VGG19 model achieved a high accuracy of 94%, demonstrating its effectiveness in capturing relevant features for distinguishing between diabetic retinopathy, glaucoma, and cataract eye diseases. The proposed system results in an increase in early detection rates, ultimately leading to better patient management.