Teleophthalmology holds a great potential to improve the quality, access, and affordability in health care. For patients, it can reduce the need for travel and provide the access to a superspecialist. Ophthalmology lends itself easily to telemedicine as it is a largely image based diagnosis. The main goal of the proposed system is to diagnose the type of disease in the retina and to automatically detect and segment retinal diseases without human supervision or interaction. The proposed system will diagnose the disease present in the retina using a neural network based classifier.The extent of the disease spread in the retina can be identified by extracting the textural features of the retina. This system will diagnose the following type of diseases: Diabetic Retinopathy and Drusen.