群脑协作的协同式脑-机接口研究进展  

Research progress of collaborative brain-computer interface for group brain collaboration

作　　者：张力新[1,2] 陈小翠 顾斌 陈龙[2] ZHANG Lixin;CHEN Xiaocui;GU Bin;CHEN Long(School of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin 300072;Academy of Medical Engineering and Translational Medicine, Tianjin University, Tianjin 300072)

机构地区：[1]天津大学精密仪器与光电子工程学院,天津300072 [2]天津大学医学工程与转化医学研究院,天津300072

出　　处：《北京生物医学工程》2020年第5期535-541,共7页Beijing Biomedical Engineering

基　　金：国家重点研发计划项目(2017YFB1002504);国家自然科学基金重点项目(81630051);天津市科技计划项目(17ZXRGGX00020、17ZXRGGX00010、16ZXHLSY00270)资助。

摘　　要：传统的非侵入型脑-机接口(brain-computer interface,BCI)系统通常采用单人-单机架构,其信息传输速率较低且鲁棒性差,难以满足高精度、多指令、短时限等复杂作业的性能需求。随着传感和信息技术的迅速发展,面向多人-多机的协同式脑-机接口系统(collaborative BCI,cBCI)应运而生。cBCI可充分发挥群体智慧优势,深入挖掘群体神经响应信息,从而更高效地完成人-机交互作业。本文综述了cBCI的基本系统架构,并结合现有研究分析其在决策与控制两个应用场景下的作业特点,讨论了面向不同作业需求和系统架构的群体神经信息融合算法的优势与不足,展望了cBCI的系统优化方向与应用研究的发展趋势。Traditional noninvasive brain-computer interface system usually adopt single-user and single-machine architecture.It’s difficult to meet the requirements of high accuracy,multiple instruction and short time delay,because of the low information transmission rate and poor robustness.With the development of sensing and information technology,the collaborative brain-computer interface based on multi-users and multi-machines has emerged.In order to complete the human-machine interaction work more efficiently,cBCI could give full play to the collective intelligence and deep digging in neural response information of group.This paper reviewed the basic architecture of the cBCI system and its application to decision-making and control.It also discussed the advantages and disadvantages of fusion algorithm for group neural response information.The system optimization direction and research trend of cBCI were proposed in the end.

关 键 词：协同脑-机接口 目标检测 运动控制 特征融合 决策融合 

分 类 号：R318.04[医药卫生—生物医学工程]