Scholarlink

<p>Early detection of Mild Cognitive Impairment (MCI) is critical for timely intervention and management of neurodegenerative diseases. In this study, we propose a novel multimodal deep learning framework that fuses electroencephalography (EEG) and heart rate variability (HRV) signals recorded during motor tasks to enhance MCI detection accuracy. The proposed Hybrid Deep Learning (HyDL) model integrates Empirical Mode Decomposition (EMD) for EEG feature extraction, a CNN-LSTM architecture for modeling spatiotemporal dynamics, and a multi-head attention mechanism for feature-level fusion. Our approach demonstrates robust performance, achieving a validation accuracy of 91.59% and a weighted F1-score of 0.94, with training accuracy exceeding 96%. These findings highlight the value of motor-task-based physiological signals and attention-driven fusion for non-invasive, interpretable early-stage cognitive decline screening.</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

Empirical Mode Decomposition and Attention-Driven Deep Fusion of Electroencephalographic and Heart Rate Variability Data for Cognitive Impairment Detection

  • 期刊论文

  • 发表日期：

    2025年9月27日

查看全文
  • Basma Jalloul,
  • Siwar Chaabene,
  • Bassem Bouaziz,
  • Walid Mahdi

摘要

Early detection of Mild Cognitive Impairment (MCI) is critical for timely intervention and management of neurodegenerative diseases. In this study, we propose a novel multimodal deep learning framework that fuses electroencephalography (EEG) and heart rate variability (HRV) signals recorded during motor tasks to enhance MCI detection accuracy. The proposed Hybrid Deep Learning (HyDL) model integrates Empirical Mode Decomposition (EMD) for EEG feature extraction, a CNN-LSTM architecture for modeling spatiotemporal dynamics, and a multi-head attention mechanism for feature-level fusion. Our approach demonstrates robust performance, achieving a validation accuracy of 91.59% and a weighted F1-score of 0.94, with training accuracy exceeding 96%. These findings highlight the value of motor-task-based physiological signals and attention-driven fusion for non-invasive, interpretable early-stage cognitive decline screening.