Scholarlink

<p>Currently, the Alzheimer’s Disease (AD) is a leading cause of dementia, that is interpreted by the aggregation of amyloid beta plaques and tau protein tangles in the brain. Early detection of the AD and its effective management pose a serious challenge to the research community. Even though state-of-the-art solutions offer certain approaches, early diagnosis remains a challenge as AD is undetected until after significant cognitive decline that in turn hinders effectiveness of the available treatments. One of the proposed solutions to tackle this issue is through the use of multimodal data i.e., through various biomarkers. In this article, we address the above issue by analyzing the integration of clinical data, and cognitive assessments scores, that helps to diagnose AD at early stages. Our proposed model utilizes Support Vector Machine (SVM) on each modality to build individual disease predictors, and implements an ensemble learning technique <i>viz.</i>, Random Forest (RF) that combines the outputs of these individual SVMs into the proposed model. We show that proposed model achieves an accuracy of 86.82%, and results in curating personalized treatment strategies by enabling timely intervention and early diagnosis.</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

Comparative Performance Analysis of ML Models for AD Diagnosis: Using Clinical and Cognitive Assessment Data

  • 期刊论文

  • 发表日期：

    2025年2月27日

查看全文
  • Manasi Maridevarmath,
  • Shreya Dugani,
  • Uddhav Mense,
  • Santosh Pattar,
  • Sridhar Iyer

摘要

Currently, the Alzheimer’s Disease (AD) is a leading cause of dementia, that is interpreted by the aggregation of amyloid beta plaques and tau protein tangles in the brain. Early detection of the AD and its effective management pose a serious challenge to the research community. Even though state-of-the-art solutions offer certain approaches, early diagnosis remains a challenge as AD is undetected until after significant cognitive decline that in turn hinders effectiveness of the available treatments. One of the proposed solutions to tackle this issue is through the use of multimodal data i.e., through various biomarkers. In this article, we address the above issue by analyzing the integration of clinical data, and cognitive assessments scores, that helps to diagnose AD at early stages. Our proposed model utilizes Support Vector Machine (SVM) on each modality to build individual disease predictors, and implements an ensemble learning technique viz., Random Forest (RF) that combines the outputs of these individual SVMs into the proposed model. We show that proposed model achieves an accuracy of 86.82%, and results in curating personalized treatment strategies by enabling timely intervention and early diagnosis.