基于脑信号的直接脑机交互(Brain-computer interaction,BCI)是一种新型的人机交互技术,可望为严重运动残疾人重建运动控制,具有潜在的医学应用价值。然而，传统的运动想象范式BCI仅识别想象运动的肢体类型，不提供运动的参数，只提供少量的方向控制命令，难于满足灵活的运动控制。为此，本课题提出一种新的运动速度和握力变化想象的BCI范式，以探索解决上述问题的途径。课题聚焦研究运动速度和握力变化想象对脑信号的调制机理。采用时间分辨高的脑电(EEG)和空间分辨率高的功能磁共振成像(fMRI)相结合的方法，揭示运动速度和握力变化想象的神经振荡、运动相关皮层电位以及脑网络的激活分布规律。最后，利用这些规律，引入先进的人机协同自适应算法和最优化的神经反馈技术，构建一个脑控服务机器人灵活运动的原型系统。该研究能够为该类BCI的传感器定位、特征提取和模式分类打下坚实的基础。

Direct brain-computer interaction(BCI) based on brain signals is a new human-computer interaction technology which is expected to reconstruct movement control for the severe motor disabled and has potential medical applications.However,the BCI based on traditional motory imagery paradigm only recognizes limb types involving imagined movement and does not provide the motion parameters and only a few direction control commands.It is difficult to meet the needs of flexible motion control.Therefore, this project will propose a new BCI paradigm of imagined movement speed and imagined variation in grip force in order to find ways for solving the aboved problem.The study will focus on the mechanism of imagined movement speed and imagined variation in grip force regulating brain signal.The activation distribution of brain networks of the neural oscillations, movement related cortical potential and BOLD is revealed by EEG with a high time resolution combined with functional magnetic resonance imaging (fMRI))with a high spatial resolution.Finally,using these mechanisms, a prototype system of brain-controlled service robot flexible movement is constructed by the introduction of the advanced man-machine Coordinated Adaptive Algorithm and optimized neurofeedback technology.This study will lay a solid foundation for the location of the sensor, feature extraction and pattern classification for the BCI.