prediction and classification of diabetic retinopathy using deep learning techniques

Abstract

newline The prevalence of DR, a primary cause of blindness, is rising every year. For the purpose of preventing vision loss, prompt and quick diagnosis of the disease is essential. It takes long time to manually diagnose DR at earliest stage, qualified experts were required to evaluate retinal fundus images. Diabetic retinopathy is primarily diagnosed by ophthalmologists who detect lesions connected to vascular irregularities based on complications from diabetes. In remote places, missing follow-ups, communication failures, and treatment delays are frequently encountered by medical experts in evaluating the fundus images. Hence the development of automated DR screening technologies is in urgency. For this, a number of techniques have been investigated, including image classification, machine learning (ML), and deep learning (DL).