Scholarlink

With the global aging population, Alzheimer’s Disease (AD) increasingly threatens the well-being of the elderly, but there’s no effective treatment yet. This study assesses the performance of a ResNet50 model enhanced with the Efficient Channel Attention (ECA) mechanism for AD diagnosis. By integrating the ECA module after the second stage of ResNet50, our research achieves efficient feature extraction and precise classification of AD imaging data. Utilizing the Kaggle Alzheimer’s Classification Dataset and Preprocessed Alzheimer Disease MRI Dataset for validation, the proposed model demonstrated accuracies of 92.50% and 93.36%, respectively. The model’s significant performance improvements across all diagnostic groups were confirmed on multiple public medical imaging datasets, establishing its efficacy and practicality as a diagnostic tool for AD.

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

Classification of Alzheimer’s Disease Using Efficient Channel Attention Mechanism

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Tao Qian,
  • Zezhou Hao,
  • Xufeng Yao

摘要

With the global aging population, Alzheimer’s Disease (AD) increasingly threatens the well-being of the elderly, but there’s no effective treatment yet. This study assesses the performance of a ResNet50 model enhanced with the Efficient Channel Attention (ECA) mechanism for AD diagnosis. By integrating the ECA module after the second stage of ResNet50, our research achieves efficient feature extraction and precise classification of AD imaging data. Utilizing the Kaggle Alzheimer’s Classification Dataset and Preprocessed Alzheimer Disease MRI Dataset for validation, the proposed model demonstrated accuracies of 92.50% and 93.36%, respectively. The model’s significant performance improvements across all diagnostic groups were confirmed on multiple public medical imaging datasets, establishing its efficacy and practicality as a diagnostic tool for AD.