Scholarlink

<p>The medical field now has become data-driven and enforces the use of machine learning to handle a large amount of data. Dementia is a critical condition of the brain that affects its basic functionality, one such disease is Alzheimer’s. Alzheimer’s Disease (AD) leads to memory loss and at the critical stage may result in the death of the patient. AD can be detected by using Classification and Content-Based Medical Image Retrieval (CBMIR) approaches and this task is accomplished by utilizing the same CNN Model. Also, the developed system has used a pipeline approach for obtaining a trained model and a parallel approach for classification and image retrieval simultaneously. The developed system has achieved 99.8% accuracy for classification with an Average Precision and Recall of retrieval system of 99.83% and 99.86%. The comparison with existing work shows that the developed model performs approximately 5% better than the existing models.</p>

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

Convolution Neural Network-Based Alzheimer Disease Detection System Using Medical Image Retrieval Approach with Multi-Class Classification

  • 期刊论文

  • 发表日期：

    2025年6月25日

查看全文
  • K. V. Sudheesh,
  • Kiran Puttegowda,
  • H. N. Naveenkumar,
  • K. Chethan,
  • Mahadevaswamy

摘要

The medical field now has become data-driven and enforces the use of machine learning to handle a large amount of data. Dementia is a critical condition of the brain that affects its basic functionality, one such disease is Alzheimer’s. Alzheimer’s Disease (AD) leads to memory loss and at the critical stage may result in the death of the patient. AD can be detected by using Classification and Content-Based Medical Image Retrieval (CBMIR) approaches and this task is accomplished by utilizing the same CNN Model. Also, the developed system has used a pipeline approach for obtaining a trained model and a parallel approach for classification and image retrieval simultaneously. The developed system has achieved 99.8% accuracy for classification with an Average Precision and Recall of retrieval system of 99.83% and 99.86%. The comparison with existing work shows that the developed model performs approximately 5% better than the existing models.