Detection of Severity Level of Diabetic Retinopathy in Retinal Fundus Images by Rule Based Fusion Method using Deep Learning Approach

Abstract

Diabetic Retinopathy (DR) is an eye disorder caused by retinal damage induced by the long-term illness of diabetes mellitus. DR is classified into two types based on the severity of the disease. The first is Non-Proliferative Diabetic Retinopathy (NPDR), which includes Normal, Mild, Moderate, and Severe classes, and the second is Proliferative Diabetic Retinopathy (PDR), which refers to the formation of abnormal blood vessels in the retina and is usually considered as the last stage. Diabetic Retinopathy may initially be asymptomatic but it can eventually result in severe visual impairment. The eye affected with DR would have features like hemorrhages, microaneurysms, exudates, and abnormal growth of blood vessels. This can be controlled or averted if detected in time, and preventive measures are taken to stop any further damage to the retina newline