Abstract:
Human vision is formed at the innermost layer of eye called retina which contains light sensitive cells namely rods and cones to discriminate intensity and color information respectively. Retina is composed of many layers. Due to excessive hyperglycemia, fluid and protein deposits are accumulated within retinal layers. This accumulation causes the vision to become distorted and blurry. If these abnormalities are not identified in early stages, it leads to severe visual impairments or blindness. Optical coherence tomography (OCT) is the best and most widely used eye testing techniquebecause of its ability to give accurate visualization of underlying pathology. Many researchers have worked on the automated detection of retinal diseases from OCT imagery but to the best of our knowledge, there is no hardware based clinical decision support system that can predict various pathological conditions of human retina. Therefore, we present a first ever hardware based hierarchical clinical decision support system that automatically screens healthy and diseased retinal OCT scans. The diseased scans are further diagnosed as retinal edema or central serous retinopathy positive based upon the type and morphology of the retinal fluid. The proposed system has been validated on random OCT scans which are publicly available online and produce the accuracy, sensitivity and specificity ratings of 85%, 80% and 90% respectively.
