A computational approach to predict diabetic retinopathy through data analytics

Abstract

Making use of estimating methods in the field of medicine has been the powerful research recently. Diabetic retinopathy is a retinal disease which causes huge blindness. Recurrent screening for prior disease detection has been a highly labor force—and resource—powerful process. So computerized diagnosis of these diseases through estimating methods would be a great remedy. Through this paper, a novel estimation strategy for computerized disease prognosis is suggested, which utilizes retinal image analysis and mining methods to accurately differentiate between the retinal images as normal and affected. Eighteen feature relevance and three variations algorithms were analyzed and used to identify the contributing features that provided better conjecture results.