Scholarlink

In every part of the world, Alzheimer’s disease (AD) represents a significant hurdle for healthcare systems due to its progressive cognitive decline and memory loss. In order to effectively intervene and manage AD, an early diagnosis and accurate diagnosis are essential. Personalized treatment strategies are being enabled by recent advancements in medical imaging and machine learning. This study uses MRI scans to classify AD using a Fusion Neural Network Model. SMOTE (Synthetic Minority Over-sampling Technique) is used to preprocess and address class imbalances in the dataset to ensure that the models perform robustly at all severity levels. Feature extraction is performed with a Convolutional Neural Network (CNN), and sequence modeling is performed with a Recurrent Neural Network (RNN). CNN extracts spatial features from MRI images, while the RNN captures temporal dependencies within sequences of feature maps. The experimental results highlight the efficacy of the Fusion Neural Network Model, achieving a test accuracy of 98.52% across AD severity stages.

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

Multi-class Classification of Alzheimer Disease Using Fusion Neural Network Model

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • B. Gunasundari,
  • K. Valli Priyadharshini,
  • Rondla Prapulla Kumar,
  • K. Guna Gayathri Praseetha,
  • S. N. Manoharan,
  • S. Sathya

摘要

In every part of the world, Alzheimer’s disease (AD) represents a significant hurdle for healthcare systems due to its progressive cognitive decline and memory loss. In order to effectively intervene and manage AD, an early diagnosis and accurate diagnosis are essential. Personalized treatment strategies are being enabled by recent advancements in medical imaging and machine learning. This study uses MRI scans to classify AD using a Fusion Neural Network Model. SMOTE (Synthetic Minority Over-sampling Technique) is used to preprocess and address class imbalances in the dataset to ensure that the models perform robustly at all severity levels. Feature extraction is performed with a Convolutional Neural Network (CNN), and sequence modeling is performed with a Recurrent Neural Network (RNN). CNN extracts spatial features from MRI images, while the RNN captures temporal dependencies within sequences of feature maps. The experimental results highlight the efficacy of the Fusion Neural Network Model, achieving a test accuracy of 98.52% across AD severity stages.