• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Circuits, Systems and Simulation >An Integrated Fall Prevention System with Single-Channel EEG and EMG Sensor
【24h】

An Integrated Fall Prevention System with Single-Channel EEG and EMG Sensor

机译：具有单通道EEG和EMG传感器的集成坠落系统

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Falls are the leading cause of injury and death in the global population, it is a major concern for home-alone elderly. The proposed system aims to provide a low-cost EEG and EMG-based fall prevention system by providing real-time feedback to the user. At the front end, a single EEG electrode placed at Cz position is used to detect movement intention by extracting the Movement Related Cortical Potential (MRCP) patterns in real-time, experimental results showed 83.3% accuracy. Upon the detection of MRCP, the system will send a cue to the EMG sensor, and the user’s EMG reading will be analysed and compared against the pre-characterized user profile. A match filter is used to determine whether the subject is fit to move or should stop moving. This process will be completed within 500ms and would leave enough time to alert the user should the system determine that the muscle state of the user is weaker than normal since MRCP typically occurs 1.5 seconds before the actual movement. The EMG match filter has an accuracy of 80%. In addition, a real-time posture detection utilizing supervised machine learning was also incorporated to realize continuous monitoring for unbalanced conditions when the user is moving. For classification, the K-nearest neighbours (KNN) classifier was chosen for a good accuracy result. It is able to identify stationary, walking, and unbalanced walking states. The accuracy rate for the KNN model is 89%.
机译：瀑布是全球人口受伤和死亡的主要原因，这是对自制老年人的重大关注。建议的系统旨在通过向用户提供实时反馈来提供低成本的EEG和基于EMG的堕落系统。在前端，置于CZ位置的单个EEG电极通过实时提取运动相关皮质电位（MRCP）图案来检测运动意图，精度显示为83.3％。在检测到MRCP时，系统将向EMG传感器发送提示，并将分析用户的EMG读数并与预先表征的用户简档进行比较。匹配过滤器用于确定主题是否适合移动或应停止移动。这个过程将在500毫秒内完成，并且会留下足够的时间来提醒用户，如果系统确定用户的肌肉状态较弱，因为MRCP通常发生在实际运动之前1.5秒。 EMG匹配过滤器的精度为80％。此外，还结合了利用监督机学习的实时姿势检测，以实现用户移动时的不平衡条件的连续监测。对于分类，选择K-CORMALY邻居（KNN）分类器以获得良好的精度结果。它能够识别静止，走路和不平衡的行走状态。 KNN模型的准确率为89％。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号