一种基于两种不同范式的混合型脑-机接口系统

混合模式脑机接口是脑-机接口研究的一个新方向,它为进一步提高脑-机接口系统性能提供了可能。现有的混合模式脑-机接口所采用的范式通常需要借助较强的视觉刺激,容易引起受试者疲劳等问题。本研究提出将运动想象和运动起始时刻视觉诱发电位两种无需强烈视觉刺激的范式以串行的方式相结合,通过运动起始时刻视觉诱发电位控制字符的输入,通过运动想象控制界面的开关和允许输入下一字符,实现了一种可用于字符输入的混合模式脑-机接口系统。为了验证系统的可行性,共完成了5例实验。实验中受试者首先进行两种范式的训练,然后进行开关系统界面和输入字符的测试。实验结果显示,经过一定训练的受试可以较好地完成系统的操作,运动想象单步操作平均时间最短为3.9 s,字符输入的正确率最高可达93.3%。除了不容易令受试者产生疲劳外,本系统相比单一感觉模式的脑-机接口也具有可完成任务种类多、控制方式灵活等优势。

A Hybrid Brain-Computer Interface System Based on Two Different Paradigms

The hybrid brain-computer interface is a new research direction in brain-computer interface which provides a possibility to further enhance the performance of brain-computer interface system.Many of existing hybrid brain-computer interfac uses paradigms with strong visual stimulation,which is likely lead to subjects’ fatigue.In this study,a hybrid brain-computer interface system for character input was proposed.The system combined motor imagery with motion-onset visual evoked potential paradigms sequentially.In this system,motion-onset visual evoked potential was used for character input,while motor imagery was used to switch the interface and allow to input the next character.Both these two paradigms do not depend on strong visual stimulations.In order to verify the feasibility of the system,a total of 5 cases of experiments have been completed.In the experiments,subjects first performed training of the two paradigms,then accepted tests of switching the system interface and input characters.Results of the experiments show that subjects can operate the system well after certain training,with a shortest average time of 3.9 s in single-step motor imagery operation,and accuracy up to 93.3% in character input.The system provides less fatigue,more task types and more flexible control than the single sensory modality based brain-computer interfaces.