Retinal Image Analysis for Abnormality Detection-An Overview

Abstract

Problem statement: Classification plays a major role in retinal image analysis for detecting the various abnormalities in retinal images. Classification refers to one of the mining concepts using supervised or unsupervised learning techniques. Approach: Diabetic retinopathy is one of the common complications of diabetes. Unfortunately, in many cases, the patient is not aware any symptoms until it is too late for effective treatment. Diabetic retinopathy is the leading cause of blindness. Diabetic retinopathy results in retinal disorders that include microaneursyms, drusens, hard exudates and intra-retinal micro-vascular abnormalities. Results: An automatic method to detect various lesions associated with diabetic retinopathy facilitate the opthalmologists in accurate diagnosis and treatment planning. Abnormal retinal images fall into four different classes namely Non-Proliferative Diabetic Retinopathy (NPDR), Central Retinal Vein Occlusion (CRVO), Choroidal Neo-Vascularization Membrane (CNVM) and Central Serous Retinopathy (CSR). Conclusion: In this study, we have analyzed the various methodologies for detecting the abnormalities in retinal images automatically along with their merits and demerits and proposed the new framework for detection of abnormalities using Cellular Neural Network (CNN).