Development of brain-computer interface for motor and verbal functions based on signal analysis of electrocorticogram

脑机接口（brain -computer interface， BCI）是指通过计算机对大脑活动的信号进行解码。我们通过分析与暂时植入硬脑膜下网格电极的神经外科患者合作获得的执行运动或言语任务的皮质电图，开发了脑机接口。结果表明，在运动功能（三种不同的手部动作）方面，我们获得了86.6±5.8%的区别表现，在分类表现方面，我们获得了53-63%的机会表现，这大大高于言语强度（默读三个元音）方面的机会水平（33%）。

Brain-computer interface(BCI)implies decoding the (human) intension from the brain activity by means of computer. We have developed BCI by analyzing electrocorticograms obtained performing either motor or verbal tasks with the cooperation of neurosurgical patients who have been temporarily implanted subdural grid electrodes. As a result, we obtained 86.6±5.8% of discriminating performance concerning motor function (three different hand movements) and 53-63% of classifying performance, which is much higher than the chance level (33%) concerning verbal intension (silent reading of three vowels).