Scholarlink

<p>Primary cause of dementia in older persons is Alzheimer’s disease (AD). Neurodegenerative alterations impact the brain in Alzheimer's disease. When diagnosing Alzheimer's disease, neurologists must manually review brain scans and perform cognitive evaluations in order to accurately identify the illness’s symptoms and trajectory. It has been discovered that human vision is not capable of recognizing subtle changes in underlying brain structure that may contain significant data about a patient's disease state. This is because subtle changes in brain anatomy can be noticed years before distinct biomarkers. So, in this proposed approach a transfer learning (TL) based deep learning (DL) approach with meta-heuristic optimization is developed for predicting alzheimer disease. The input image is initially pre-processed using Wavelet Structure using Self-guidance (WSSG) and Low light image (LLI) techniques for eliminating noise and enhance the quality of a raw image. Then, the image is seperated by SwinE-Net technique. Finally classification is done by using Transfer learning based Deep learning approach termed as CATNet with secretary bird optimization (SBO) for predicting the alzheimer disease. SBO optimization is used to select the hyper parameters of CATNet optimally. The proposed method achieves performance such as accuracy of 98.2%, precision of 96.16% and selectivity of 98.74%. Thus, the proposed segmentation and TL based optimized DL is a better choice for predicting Alzheimer stages accurately.</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

Alzheimer stages prediction using swinnet for segmentation and transfer learning based CATNet approach with SBO algorithm

  • 期刊论文

  • 发表日期：

    2025年1月29日

查看全文
  • Bhuvaneshwari Jolad,
  • M. Venkateswara Rao,
  • Subhash Y. Kamdi,
  • Rahul N. Patil,
  • Lakshmana Phaneendra Maguluri

摘要

Primary cause of dementia in older persons is Alzheimer’s disease (AD). Neurodegenerative alterations impact the brain in Alzheimer's disease. When diagnosing Alzheimer's disease, neurologists must manually review brain scans and perform cognitive evaluations in order to accurately identify the illness’s symptoms and trajectory. It has been discovered that human vision is not capable of recognizing subtle changes in underlying brain structure that may contain significant data about a patient's disease state. This is because subtle changes in brain anatomy can be noticed years before distinct biomarkers. So, in this proposed approach a transfer learning (TL) based deep learning (DL) approach with meta-heuristic optimization is developed for predicting alzheimer disease. The input image is initially pre-processed using Wavelet Structure using Self-guidance (WSSG) and Low light image (LLI) techniques for eliminating noise and enhance the quality of a raw image. Then, the image is seperated by SwinE-Net technique. Finally classification is done by using Transfer learning based Deep learning approach termed as CATNet with secretary bird optimization (SBO) for predicting the alzheimer disease. SBO optimization is used to select the hyper parameters of CATNet optimally. The proposed method achieves performance such as accuracy of 98.2%, precision of 96.16% and selectivity of 98.74%. Thus, the proposed segmentation and TL based optimized DL is a better choice for predicting Alzheimer stages accurately.