Scholarlink

Background <p>The rising prevalence of neurodegenerative disorders, particularly Alzheimer’s disease (AD) and frontotemporal dementia (FTD), poses an escalating healthcare challenge worldwide. Electroencephalography (EEG) provides a promising approach for investigating underlying neural mechanisms, yet studies have shown inconsistent findings. This study implemented a comprehensive analytical framework combining spectral, nonlinear dynamics, and graph theoretical approaches to characterize EEG patterns in AD and FTD.</p> Methods <p>We analyzed EEG recordings from 36 AD patients, 23 FTD patients, and 29 healthy controls (HC), and established machine learning models with model performance evaluated using classification accuracy and area under the receiver operating characteristic curve (AUC).</p> Results <p>Group-level analyses with cluster-based correction revealed distinct and frequency-dependent EEG alterations between AD and FTD. AD was characterized by more pronounced posterior abnormalities, including increased theta activity and reduced alpha- and beta-band power, whereas FTD showed relatively intermediate changes with a more central distribution. Nonlinear dynamics analyses further indicated disease-specific alterations in signal complexity across frequency bands. Graph theoretical analysis demonstrated distinct patterns of disrupted brain organization between the two conditions. In addition, machine learning results indicated that graph theoretical measures achieved the highest classification performance in distinguishing AD from FTD, with an accuracy of 81.36%.</p> Conclusions <p>These findings delineate distinct neurophysiological profiles of AD and FTD across multiple analytical dimensions and support the relevance of graph theoretical analysis for differentiating dementia subtypes, providing a basis for further investigation of the neural mechanisms underlying AD–FTD differentiation.</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

Multi-dimensional EEG analysis reveals distinct neurophysiological patterns in Alzheimer’s and frontotemporal dementia

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月1日

查看全文
  • Guiyuan Cai,
  • Yu Shi,
  • Junqin Ma,
  • Xuefei Zhang,
  • Wen Wu

摘要

Background

The rising prevalence of neurodegenerative disorders, particularly Alzheimer’s disease (AD) and frontotemporal dementia (FTD), poses an escalating healthcare challenge worldwide. Electroencephalography (EEG) provides a promising approach for investigating underlying neural mechanisms, yet studies have shown inconsistent findings. This study implemented a comprehensive analytical framework combining spectral, nonlinear dynamics, and graph theoretical approaches to characterize EEG patterns in AD and FTD.

Methods

We analyzed EEG recordings from 36 AD patients, 23 FTD patients, and 29 healthy controls (HC), and established machine learning models with model performance evaluated using classification accuracy and area under the receiver operating characteristic curve (AUC).

Results

Group-level analyses with cluster-based correction revealed distinct and frequency-dependent EEG alterations between AD and FTD. AD was characterized by more pronounced posterior abnormalities, including increased theta activity and reduced alpha- and beta-band power, whereas FTD showed relatively intermediate changes with a more central distribution. Nonlinear dynamics analyses further indicated disease-specific alterations in signal complexity across frequency bands. Graph theoretical analysis demonstrated distinct patterns of disrupted brain organization between the two conditions. In addition, machine learning results indicated that graph theoretical measures achieved the highest classification performance in distinguishing AD from FTD, with an accuracy of 81.36%.

Conclusions

These findings delineate distinct neurophysiological profiles of AD and FTD across multiple analytical dimensions and support the relevance of graph theoretical analysis for differentiating dementia subtypes, providing a basis for further investigation of the neural mechanisms underlying AD–FTD differentiation.