Scholarlink

To find a treatment that works and avoids severe vision loss, Diabetic Retinopathy (DR) must be effectively detected and graded. However, DR screening programmers face difficulties due to the accessibility of human graders and the inefficiency of manual fundus image analysis. The research presented here introduces a comprehensive approach that combines both Machine Learning (ML) and Deep Learning (DL) approaches for enhancing the advanced detection of diabetic retinopathy (DR) and screening diabetics' eyes for damage. Convolutional Neural Networks (CNNs) with Grey Wolf Optimization (GWO) and classifiers like Support Vector Machines (SVM) and XGBoost can be combined in an innovative methodology called the CGSX ensemble. Our proposed work has the best overall performance, 0.95 F1 score achieving 0.95 precision, 0.94 recall, and 0.94 accuracy.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An Ensemble Approaches to Improve Diabetic Retinopathy (DR) Detection and Severity Classification from Retinal Fundus Images

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • K. Kayathri,
  • K. Kavitha

摘要

To find a treatment that works and avoids severe vision loss, Diabetic Retinopathy (DR) must be effectively detected and graded. However, DR screening programmers face difficulties due to the accessibility of human graders and the inefficiency of manual fundus image analysis. The research presented here introduces a comprehensive approach that combines both Machine Learning (ML) and Deep Learning (DL) approaches for enhancing the advanced detection of diabetic retinopathy (DR) and screening diabetics' eyes for damage. Convolutional Neural Networks (CNNs) with Grey Wolf Optimization (GWO) and classifiers like Support Vector Machines (SVM) and XGBoost can be combined in an innovative methodology called the CGSX ensemble. Our proposed work has the best overall performance, 0.95 F1 score achieving 0.95 precision, 0.94 recall, and 0.94 accuracy.