Scholarlink

Diabetic Macular Edema (DME) is a critical complication of diabetes mellitus and is responsible for vision impairment and blindness. Due to an increase in diabetes cases, early and accurate DME detection has become important. This paper presents a deep learning-based solution for DME detection using ResNet50 and ResNet101 architectures. The proposed ResNet50 model achieved an accuracy of 94.56%, sensitivity of 95.98%, specificity of 87.8%, and precision of 97.45%, surpassing existing methods. The ResNet101 model, while slightly lower in accuracy at 90.83%, provides a high sensitivity of 91.83% and precision of 97.45%, with a specificity of 84.38%. This research demonstrates the transformative potential of deep learning in revolutionizing DME diagnosis.

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

A Residual Network-Based Approach for Early Diabetic Macular Edema Detection

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Nitigya Sambyal,
  • Sukhpal Singh

摘要

Diabetic Macular Edema (DME) is a critical complication of diabetes mellitus and is responsible for vision impairment and blindness. Due to an increase in diabetes cases, early and accurate DME detection has become important. This paper presents a deep learning-based solution for DME detection using ResNet50 and ResNet101 architectures. The proposed ResNet50 model achieved an accuracy of 94.56%, sensitivity of 95.98%, specificity of 87.8%, and precision of 97.45%, surpassing existing methods. The ResNet101 model, while slightly lower in accuracy at 90.83%, provides a high sensitivity of 91.83% and precision of 97.45%, with a specificity of 84.38%. This research demonstrates the transformative potential of deep learning in revolutionizing DME diagnosis.