Scholarlink

Dementia-related diseases, such as Alzheimer’s, require early detection to slow cognitive decline and improve patient care. Electroencephalography (EEG) is a non-invasive method that can detect abnormal brain activity associated with these conditions. In this study, we developed an IoMT-based system using the STM32 microcontroller to enable real-time, embedded detection of dementia episodes. Using an LSTM model that achieved the highest accuracy (98%), surpassing SVM (95%) and Random Forest (75%). The system demonstrated an efficient execution time of 35 s per EEG signal, proving the feasibility of real-time edge computing for dementia monitoring.

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

Embedded IoMT System for Early Dementia Detection Using EEG Signals and Deep Learning on STM32 Microcontroller

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • F. Jakjoud

摘要

Dementia-related diseases, such as Alzheimer’s, require early detection to slow cognitive decline and improve patient care. Electroencephalography (EEG) is a non-invasive method that can detect abnormal brain activity associated with these conditions. In this study, we developed an IoMT-based system using the STM32 microcontroller to enable real-time, embedded detection of dementia episodes. Using an LSTM model that achieved the highest accuracy (98%), surpassing SVM (95%) and Random Forest (75%). The system demonstrated an efficient execution time of 35 s per EEG signal, proving the feasibility of real-time edge computing for dementia monitoring.