A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal Images

Diabetic Retinopathy (DR) is a major cause of visual impairment worldwide, especially in diabetic patients. Early detection is extremely important for effective management and the prevention of serious complications. The deep learning approach for classifying diabetic retinopathy using retinal imagi...

Full description

Saved in:
Bibliographic Details
Main Authors: Rathod-Jadhav Kavita, Pande Aparna
Format: Article
Language:English
Published: EDP Sciences 2025-01-01
Series:EPJ Web of Conferences
Online Access:https://www.epj-conferences.org/articles/epjconf/pdf/2025/13/epjconf_icetsf2025_01009.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849687988662960128
author Rathod-Jadhav Kavita
Pande Aparna
author_facet Rathod-Jadhav Kavita
Pande Aparna
author_sort Rathod-Jadhav Kavita
collection DOAJ
description Diabetic Retinopathy (DR) is a major cause of visual impairment worldwide, especially in diabetic patients. Early detection is extremely important for effective management and the prevention of serious complications. The deep learning approach for classifying diabetic retinopathy using retinal imaging is an automated system for classifying diabetic retinopathy using deep learning techniques. The folding neural network (CNN) model architecture analyzes retinal images and classifies them with high accuracy as diabetic or non-diabetic classification. The proposed framework includes extensive prepossessing steps such as image size, normalization, and data expansion to improve model robustness. A custom CNN model has been developed. This included a folding layer with batch normalization and dropouts to improve characteristic extraction, simultaneously reducing overhang. This model was trained on a diverse dataset of retinal images and validated using standard metrics including accuracy, accuracy, recall, F1 scores, and confusion matrix. The results show that the system achieves a high level of classification accuracy. This demonstrates its potential as a reliable tool for screening for early diabetic retinopathy. Performance analysis and training history visualization continue to validate the validity and generalizability of the model. It will help contribute in the field of medical imaging by providing an accessible, AI-controlled solution for preventive diabetic health care and highlighting the importance of early diagnosis to improve patient outcomes.
format Article
id doaj-art-019d0bfc162a46399202a83aa7fa3c81
institution DOAJ
issn 2100-014X
language English
publishDate 2025-01-01
publisher EDP Sciences
record_format Article
series EPJ Web of Conferences
spelling doaj-art-019d0bfc162a46399202a83aa7fa3c812025-08-20T03:22:11ZengEDP SciencesEPJ Web of Conferences2100-014X2025-01-013280100910.1051/epjconf/202532801009epjconf_icetsf2025_01009A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal ImagesRathod-Jadhav Kavita0Pande Aparna1Computer Science and Engineering, SunRise UniversityComputer Science and Engineering, Nutan College of Engineering and ResearchDiabetic Retinopathy (DR) is a major cause of visual impairment worldwide, especially in diabetic patients. Early detection is extremely important for effective management and the prevention of serious complications. The deep learning approach for classifying diabetic retinopathy using retinal imaging is an automated system for classifying diabetic retinopathy using deep learning techniques. The folding neural network (CNN) model architecture analyzes retinal images and classifies them with high accuracy as diabetic or non-diabetic classification. The proposed framework includes extensive prepossessing steps such as image size, normalization, and data expansion to improve model robustness. A custom CNN model has been developed. This included a folding layer with batch normalization and dropouts to improve characteristic extraction, simultaneously reducing overhang. This model was trained on a diverse dataset of retinal images and validated using standard metrics including accuracy, accuracy, recall, F1 scores, and confusion matrix. The results show that the system achieves a high level of classification accuracy. This demonstrates its potential as a reliable tool for screening for early diabetic retinopathy. Performance analysis and training history visualization continue to validate the validity and generalizability of the model. It will help contribute in the field of medical imaging by providing an accessible, AI-controlled solution for preventive diabetic health care and highlighting the importance of early diagnosis to improve patient outcomes.https://www.epj-conferences.org/articles/epjconf/pdf/2025/13/epjconf_icetsf2025_01009.pdf
spellingShingle Rathod-Jadhav Kavita
Pande Aparna
A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal Images
EPJ Web of Conferences
title A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal Images
title_full A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal Images
title_fullStr A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal Images
title_full_unstemmed A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal Images
title_short A Deep Learning Approach for Diabetic Retinopathy Classification Using Retinal Images
title_sort deep learning approach for diabetic retinopathy classification using retinal images
url https://www.epj-conferences.org/articles/epjconf/pdf/2025/13/epjconf_icetsf2025_01009.pdf
work_keys_str_mv AT rathodjadhavkavita adeeplearningapproachfordiabeticretinopathyclassificationusingretinalimages
AT pandeaparna adeeplearningapproachfordiabeticretinopathyclassificationusingretinalimages
AT rathodjadhavkavita deeplearningapproachfordiabeticretinopathyclassificationusingretinalimages
AT pandeaparna deeplearningapproachfordiabeticretinopathyclassificationusingretinalimages