Detection of Diabetic Retinopathy using Deep Learning: A Review
Amnaya Pradhan1, Neha Sharma2

1Amnaya Pradhan, Department of Computer Science Engineering, S.R.M. Institute of Science and Technology, Chennai (Tamil Nadu), India.
2Neha Sharma, Department of Computer Application, Panjab University, (Chandigarh), India.
Manuscript received on 29 June 2022 | Revised Manuscript received on 05 July 2022 | Manuscript Accepted on 15 July 2022 | Manuscript published on 30 July 2022 | PP: 138-143 | Volume-11 Issue-2, July 2022 | Retrieval Number: 100.1/ijrte.B71750711222 | DOI: 10.35940/ijrte.B7175.0711222
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: Throughout the globe, 1.6 million people annually fall prey to dia- betes. And an alarming total of 422 million people throughout the world have been diagnosed with diabetes, most of the contribution to this number being from low and middleincome countries. Diabetic retinopathy is the number one cause of blindness in the world. It generally affects people from ages 25 to 65. It occurs when the blood vessels present in the retina get damaged by hyper – glycemia or prevents blood from passing through the eyes. It is crucial to treat diabetic retinopathy early. If left untreated, it eventually leads to blindness. The proposed methodology is to use Convolutional Neural Networks with ResNet in order to diagnose diabetic retinopathy. Fundal cameras are used to obtain retinal images. The aim is to detect and prevent this disease, where it is challenging to perform medical tests. As per the research study, the images will be prepro- cessed, segmented, enhanced, and then the extraction of features such as micro aneurysms and hemorrhages will occur. Based on this, the disease will be clas- sified into mild, moderate, severe, or proliferative. In the future, this model may also be used to detect other conditions, such as glaucoma and macular degener- ation. 
Keywords: ResNet, Diabetes, Convolutional Neural Networks, Micro Aneurysms, Hemorrhages, Retinal Images.
Scope of the Article: Deep Learning