de Melo Gabriel Chaves, Martes Sternlicht Vitor, Forner-Cordero Arturo
Annu Int Conf IEEE Eng Med Biol Soc. 2020 Jul;2020:3027-3030. doi: 10.1109/EMBC44109.2020.9175851.
The identification of specific components in EEG signals is often key when designing EEG-based brain-computer interfaces (BCIs), and a good understanding of the factors that elicit such components can be helpful when it comes to precise, energy-efficient and time-accurate actuation of exoskeletons. CNVs (Contingent Negative Variations), ERDs or ERSs (Event-Related Desynchronizations/Synchronizations) as well as ErrPs (Error-Related Potentials) are particularly important components can be identified during motor tasks and related to specific events in a Coincident Timing (CT) task. This work investigates offline EEG signals acquired during an upper limb CT task and analyzes the task protocol with the purpose of correlating the aforementioned EEG features to movement onset. CNVs and ERD/ERS were successfully identified after averaging multiple trials, and it was further concluded that complementary information about muscle activity (via EMG) as well as video tracking of arm movement play a critical role in the synchronization of EEG components with movement onset. The framework for EEG analysis presented in this paper allows for future development of a BCI on top of this CT task capable of assessing motor learning and actuating an exoskeleton to enable faster motor rehabilitation.
在设计基于脑电图（EEG）的脑机接口（BCI）时，识别EEG信号中的特定成分通常是关键，而深入了解引发这些成分的因素，对于精确、节能且时间准确地驱动外骨骼可能会有所帮助。关联性负变（CNV）、事件相关去同步化/同步化（ERD/ERS）以及错误相关电位（ErrP）是在运动任务期间可以识别出的特别重要的成分，并且与同步定时（CT）任务中的特定事件相关。这项工作研究了上肢CT任务期间采集的离线EEG信号，并分析了任务协议，目的是将上述EEG特征与运动起始相关联。在对多次试验进行平均后成功识别出了CNV和ERD/ERS，并且进一步得出结论，关于肌肉活动（通过肌电图（EMG））的补充信息以及手臂运动的视频跟踪在EEG成分与运动起始的同步中起着关键作用。本文提出的EEG分析框架为未来在此CT任务之上开发能够评估运动学习并驱动外骨骼以实现更快运动康复的BCI奠定了基础。
