Abstract: Aiming at the shortages of traditional motion imagery(MI) paradigm and steady state visual evoked potential (SSVEP) paradigm in high training difficulty of paradigm, complex equipment integration, high cost, unwearable and so on, a brain control prosthesis method was proposed based on facial expression with the help of the wide adaptability of electroencephalography, high accuracy of recognition and the convenience to integrate in portable system.A portable brain controlled prosthesis system was constructed based on facial expression，including hardware system forming，analytical algorithm design and software developing，to realize the control technique from brow-raising to hand-opening，brow-frowning to hand-closing，left-smirking to wrist-intorsion and right-smirking to wrist-extorsion and its online experiments.The online experimental results of the portable brain controlled prosthesis system indicate that the brain control method for prosthesises based on facial expression is better than SSVEP way in removing the extra inducers，and better than motor imagery way for the average online accuracy rate of recognition and control is over 85%.

Key words: brain computer interface, electroencephalography, wearable device, expression, prosthesis control

摘要： 针对传统纯意念运动想象脑控范式和稳态视觉诱发脑控范式存在范式训练难度高、设备集成复杂及高成本、不易随身携带等问题，借助表情辅助脑电信号适应性广、识别率高和易于便携化系统集成的优势，提出了基于表情辅助的假手脑控方法；研究构建了一种基于表情辅助的假手便携可穿戴脑控系统，包括硬件选型设计与系统组建，及表情辅助脑电信号分类识别算法和假手控制算法的设计研究与软件模块开发等，实现提眉、皱眉、左撇嘴和右撇嘴4种辅助表情脑电信号的检测识别，对应完成假手的手指张开、手指闭合、手腕内旋和手腕外旋4个动作的控制，并进行在线实验验证。结果表明：基于表情辅助的假手脑控方法不仅优于传统的稳态视觉诱发脑控范式（去掉了系统中原所必须的稳态视觉诱发器），而且优于纯意念运动想象脑控范式，平均在线识别与控制准确率高于85%。

关键词: 脑机接口, 脑电信号, 可穿戴设备, 表情, 假手控制