Deep Learning based Human Eye Disease Predication System

Millions worldwide lose their eyesight yearly, normally from treatable eye diseases, if only identified early enough. What if we could find these conditions just before it is actually too late restorations of vision and transformations of lives? This is what our project is all about. We are devising a Deep Learning-based Eye Disease Prediction System which should be able to identify relevant eye diseases such as Glaucoma, Diabetic Retinopathy, Cataract, and Agerelated Macular Degeneration (AMD) from retinal images, both fast and accurately. To do this, we are putting Convolutional Neural Networks (CNNs) and Transfer Learning to work for us, so that our system could learn from thousands of retinal images to detect certain deep subtle patterns which may not be quite apparent to the human observer. We combine the domains of advanced image preprocessing, feature extraction, and classification to deliver predictions that have got to be very accurate. We didn't stop there we made it practical. This system has a webbased interface that allows easy access for doctors and patients, thus bringing in quick, noninvasive, and cost-effective eye diseases screening for all of them. With this tool, we hope to make early detection of eye diseases easier, faster, and cheaper this way saving the gift of vision for many.