International Journal of Advanced Computer Technology

https://ijact.org/index.php/ijact/issue/feed International Journal of Advanced Computer Technology 2025-02-18T19:31:47+00:00 Dr. Amrendra Singh Yadav editor@ijact.org Open Journal Systems <center> <h3 style="font-family: georgia,serif; font-size: 20px; color: #0066ff;">A Peer reviewed, Open access, Bimonthly, International Journal since 2012</h3> </center> <section class="additional_content"> ISSN : 2319-7900 | DOI : 10.18535/ijact | <a href="https://ijact.org/revista/index.php/ijact/gateway/plugin/WebFeedGatewayPlugin/rss">RSS</a> <a href="https://ijact.org/revista/index.php/ijact/about/submissions">Submission</a>  Call For Paper Volume 8 issue 2 December 2019 Submission Deadlines Volume / Issue / Month : Volume 08 Issue 2, 25 December 2019 Read More  Submit Article at : <a href="mailto:editor@ijact.org">submitpaper@ijact.org</a> International Journal of Advanced Computer Technology (IJACT) ISSN - 2319-7900 is a leading international online journal for publication of new ideas. It's a peer-reviewed, open access scholarly journal that publishes research works and review articles in all research domains (Engineering, Technology). This is an open access journal which means that all content is freely available to the Researcher /users of any institution with vision spreading and sharing knowledge. These published articles/papers are accessible to the research scholars through a wide indexing policy adopted by this online international journal. Hence, they can freely be accessed and used by everyone for the improvement of current trends in science and technology. This Journal favors and promotes online publication of papers to truly present itself as an online journal. </section> <center></center> IJACT follows publication ethics during phases of online publication inline with the guidelines and standards developed by the Committee on Publication Ethics (COPE) to avoid any malpractices by all concerned including authors, reviewers, editors and publishers. This is in accordance with the BOAI definition of open access. We are committed and promise to take this journal to greater heights. We invites authors for the guest editorship, reviewers, program for improving the journal quality and also sharing the innovative ideas. Frequency : Six issues/year Submit Article at submitpaper@ijact.org https://ijact.org/index.php/ijact/article/view/154 Automated Diabetic Retinopathy Detection Using Deep Learning: A Comparative Analysis of VGG-16 and ResNet50 2025-02-18T19:31:47+00:00 Ankit Sharma e22cseu0352@bennett.edu.in Parv Wadhwa e22cseu0561@bennett.edu.in Saksham Arora e22cseu0360@bennett.edu.in

Diabetic retinopathy (DR) is a leading cause of blindness worldwide, primarily affecting individuals with prolonged diabetes. Early detection is crucial for preventing severe vision loss, yet conventional diagnostic methods are time-intensive and require specialized expertise. This study proposes a deep learning-based automated DR classification system utilizing convolutional neural networks (CNNs), specifically VGG-16 and ResNet50 architectures. The model classifies DR into five categories: normal, mild, moderate, severe, and proliferative DR. A dataset of retinal fundus images was preprocessed and analyzed using these CNN models, with performance evaluated based on classification accuracy. The VGG-16 model achieved an accuracy of 79.99%, outperforming ResNet50, which attained 70%. The findings highlight the effectiveness of deep learning in automated DR screening, demonstrating its potential for enhancing early diagnosis and patient care. Further improvements, such as advanced preprocessing, data augmentation, and hybrid modelling, can refine the accuracy and clinical applicability of AI-driven diagnostic tools.