Investigating Opthalmic images to Diagnose Eye diseases using Deep Learning Techniques

In Ophthalmology, several eye image related diagnostic methodologies started offering meaningful insight about eye disease based on huge set of data points and morphological datasets. The common causes of various eye abnormalities are any trauma, aging and diseases, particularly diabetes. Glaucoma, retinal detachment, Cataract, macular degeneration, diabetic retinopathy, hypertensive Retinopathy, and diabetic macular edema are the most common causes of blindness all around the world. All these diseases are manually suspected and diagnosed by ophthalmologists with clinical practice. Lack of eye care centers and scarcity of trained ophthalmologists are very common in rural and remote areas in developing countries like India. Detection at early stage, followed by appropriate and timely clinical treatment for a varieties of eye diseases, can solve problem of blindness to a large extent. Ophthаlmоlоgist adopt a computerized analysis of ophthalmic imaging modalities for faster clinical screening and accurate diagnosis of these diseases. Latest deep learning models, when compared with traditional machine learning algorithms, speed up the diagnosis of disease and improve accuracy with less false positive rate. This paper investigates all latest deep learning methodologies used in automatic retinal disease diagnosis process concerning different image modalities. This paper also lists some of the most common eye diseases, limitations with the latest developments and potential challenges for future enhancements.