Diabetic Retinopathy Detection using Convolutional Neural Network

Abstract

the motive behind the Research is to design a system that automates the preliminary Diabetic Retinopathy Finding on the basis of retinal image of a patient's eye using Convolutional Neural Network. It is one of the important reasons of preventable blindness. Detecting it physically is a time-consuming and difficult process. In this, blood vessels in the retina get injured due to high blood sugar levels. Approximately of the blood vessels swell and outflow, some may even shrink, and thereby discontinuing the blood flow. Sometimes uneven new blood vessels develop on the retina. These variations can blur one’s vision. It sufferings up to 80 percent of people who have required diabetes for 20 years or extra. It can be identified into 5 stages: No DR, Slight, Reasonable, Simple, and Proliferative DR.

Keywords- Adaptive Learning, Convolutional Neural Networks, Deep neural networks, Image Processing, Image detection.