Research Article
Retina Grader – A Diabetic Retinopathy Classification System
@INPROCEEDINGS{10.4108/eai.28-4-2025.2358113, author={Jacquelin Anushya P and Lavanya M and Gokul S and Hariharan N and Hariharan L and Hari Prasanth A}, title={Retina Grader -- A Diabetic Retinopathy Classification System}, proceedings={Proceedings of the 4th International Conference on Information Technology, Civil Innovation, Science, and Management, ICITSM 2025, 28-29 April 2025, Tiruchengode, Tamil Nadu, India, Part II}, publisher={EAI}, proceedings_a={ICITSM PART II}, year={2025}, month={10}, keywords={cnn resnet-50 dr xai cad ai ml}, doi={10.4108/eai.28-4-2025.2358113} }- Jacquelin Anushya P
Lavanya M
Gokul S
Hariharan N
Hariharan L
Hari Prasanth A
Year: 2025
Retina Grader – A Diabetic Retinopathy Classification System
ICITSM PART II
EAI
DOI: 10.4108/eai.28-4-2025.2358113
Abstract
Diabetic Retinopathy (DR) is a serious complication of diabetes, being a major cause of blindness globally, necessitating early diagnosis for vision loss prevention. Conventional diagnostic approaches that utilize human examination of retinal fundus images are labor-intensive, error-prone and un- available in resource- limited regions. This study presents an AI- based approach using the deep learning ResNet-50 architecture to automate the detection and the classification of DR into five levels: Normal, Mild, Moderate, Severe, or Proliferative. Based on the Kaggle Diabetic Retinopathy Dataset, the system is equipped with data preprocessing including resizing, normalization, and augmentation to enhance the quality and balance of the data. ResNet-50 provides stronger feature extraction and classification to improve accuracy and reduce computation, collusion effects. An easy-to-use interface permits easy image upload, prediction and result visualization, and deployment in the cloud as well as offline makes it accessible. Such a system will contribute directly to global eye health by filling a gap in traditional diagnostics as a scalable, robust solution: preventing unnecessary blindness and disability.
