Ear-EEG Recording for Brain Computer Interface of Motor Task

Ear-EEG recording collects brain signals from electrodes placed in the ear canal. Compared with existing scalp-EEG, ear-EEG is more wearable and user-comfortable compared with existing scalp-EEG.In this dataset, we collected ear-EEG signals and the scalp-EEG signals when subjects were performing a left/right hand grasping motor task. We first validated the ear-EEG recordings by measuring the similarity of power ratio and channel correlation between ear-EEG and scalp-EEG signals. Then we applied EEG compact network (EEGNet) for classification of left/right-hand motor tasks using ear-EEG signals and scalp EEG signals separately. Our results showed motor task classification based on ear-EEG has a high potential for the practical BCI applications in motor task.耳电生理脑电图（Ear-EEG）记录技术通过置于耳道内的电极采集大脑信号。相较于现有头皮脑电图（scalp-EEG），耳电生理脑电图具备更优异的可穿戴性与用户佩戴舒适度。本数据集采集了受试者执行左右手抓握运动任务时的耳电生理脑电图信号与头皮脑电图信号。本研究首先通过计算耳电生理脑电图与头皮脑电图信号的功率比相似度及通道相关性，对耳电生理脑电图记录结果进行了有效性验证。随后我们分别采用脑电图紧凑网络（EEGNet），基于耳电生理脑电图信号与头皮脑电图信号，实现左右手运动任务的分类。实验结果表明，基于耳电生理脑电图的运动任务分类模型，在运动相关脑机接口（Brain-Computer Interface, BCI）的实际应用中具备极高的应用潜力。