Improving Information Transfer rate in Active BCIs
Improving Information Transfer rate in Active BCIs
复制标题
提高主动 BCI 中的信息传输率
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
V. D. Sa
中科院分区：
文献类型：
--
作者：
V. D. Sa
我们考虑的情况下，嘈杂的二进制EEG为基础的脑机接口，其中一个人试图产生两个可辨别的控制信号，但接收到的信号是嘈杂的，最佳分类边界（或解码器）是未知的或变化。在这种情况下，计算机在执行操作之前随着时间的推移积累证据是很常见的。可以给出中间反馈以告知用户当前的解码沿着。在这些条件下，我们已经表明，通过马尔可夫链分析，信息传输速率更高时，用户和计算机附加响应（中间反馈）的意义“继续/好”和“改变方向/坏”的两类噪声信号，他们产生的，而不是直接命令“左”和“右”。在本文中，我们分析了这些系统的第一步，并表明，当还没有一个计算机解释的响应响应，“右/左”的命令是最翔实的，一个系统，其中第一个“右/左”的步骤是结合未来的“好/坏”的交互式命令给出了最高的信息传输速率。最后，我们表明，这种混合的方法可以被看作是一个自然的游戏界面。
We consider the case of a noisy binary EEGbased brain-computer interface where a human attempts to generate two discriminable control signals but the received signals are noisy and the optimal classification boundary (or decoder) is not known or changing. In such situations, it is common for the computer to accumulate evidence over time before performing an action. Intermediate feedback can be given to inform the user of the current decoding along the way. Under these conditions we have shown via Markov chain analysis that the information transfer rate is higher when the user and computer attach the responsive (to the intermediate feedback) meanings of "continue/good" and "change-direction/bad" to the two classes of noisy signals they generate instead of direct commands "left" and "right". In this paper, we analyze the first step of these systems and show that when there is not yet a computer-interpreted response to respond to, the "right/left" commands are most informative and that a system where a first "right/left" step is combined with future "good/bad" interactive commands gives the highest information transfer rate. Finally we show that this hybrid approach can be seen as a natural game-like interface.
DOI：
10.1109/86.712231
发表时间：
1998-09-01
期刊：
IEEE transactions on rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society
影响因子：
--
作者：
Wolpaw, J R;Ramoser, H;Pfurtscheller, G
通讯作者：
Pfurtscheller, G