【摘要】 近年来,脑机接口日益成为人工智能领域的研究热点,被广泛应用于康复医疗、人工智能、游戏娱乐等领域。脑机接口通过采集人体脑电信号,并将其转化为计算机或其他外部设备的操作指令,可以使认知控制训练更加精准、客观。本文总结了脑机接口系统的基本组成,包含如何获取脑电信号,选取脑电信号处理方法,控制命令的输出以及对目标系统的控制。梳理和总结了基于脑电的传统脑机接口范式,包括P300脑机接口、运动想象脑机接口以及稳态视觉诱发电位脑机接口等范式的基本原理、研究现状和应用场景,并对比了各类范式的优缺点。最后分析脑机接口技术的局限和难点,并对该技术领域的发展与应用前景进行了展望。更多还原

【Abstract】 In recent years, brain-computer interfaces(BCI) have increasingly become a research hotspot in the field of artificial intelligence, and are widely used in rehabilitation medicine, game entertainment, military applications, home intelligence and other fields.The brain-computer interface can make cognitive control training more accurate and objective by collecting human electroencephalogram(EEG) signals and converting them into operating instructions for computers or other external devices. This paper reviews the basic components of the brain computer interface system, including how to obtain EEG signals, the selection of EEG signal processing methods, the output of control commands, and the control of the target system. Also, the traditional EEG-based brain-computer interface paradigm was introduced, including the basic principles, research status, and application scenarios of P300-BCI, motor imagery-BCI( MI-BCI), steady-state visual evoked potentials-BCI(SSVEP-BCI) and other paradigms, with the advantages and disadvantages of various paradigms pared. The risks and challenges faced by current research in terms of technology and ethics are proposed, and the development and application prospects of this field are prospected.更多还原