Netra

Abstract

Diabetic Retinopathy (DR) is a common complication of diabetes mellitus, which causes lesions onthe retina that affect vision. If it is not detected early, it can lead to blindness. Unfortunately, DRis not a reversible process, and treatment only sustains vision. DR early detection and treatmentcan significantly reduce the risk of vision loss. The manual diagnosis process of DR retina fundusimages by ophthalmologists is time-, effort-, and cost-consuming and prone to misdiagnosis unlikecomputer-aided diagnosis systems.[ 1] Convolutional neural networks are more widely used asa deep learning method in medical image analysis and they are highly effective.[1] Netrascopyis a more efficient system for Diabetic Retinopathy detection, which consists of a low cost,Camera, “DIYretCAM Netrascopy FUNDUS Camera V1”, An Android Application and WebApplication which aims to help patients and doctors detect diabetic retinopathy at early stages bytaking 30-Second video of patient’s retina and passing each frame as an individual test case to aConvolutional Neural Network to detect probability of a patient having diabetic retinopathy.