Key Laboratory of Advanced Control and Optimization for Chemical Processes, Ministry of Education, East China University of Science and Technology, Shanghai 200237, PR China.
J Neurosci Methods. 2012 Apr 15;205(2):265-76. doi: 10.1016/j.jneumeth.2012.01.004. Epub 2012 Jan 16.
Brain-computer interfaces (BCIs) allow users to communicate via brain activity alone. Many BCIs rely on the P300 and other event-related potentials (ERPs) that are elicited when target stimuli flash. Although there have been considerable research exploring ways to improve P300 BCIs, surprisingly little work has focused on new ways to change visual stimuli to elicit more recognizable ERPs. In this paper, we introduce a "combined" BCI based on P300 potentials and motion-onset visual evoked potentials (M-VEPs) and compare it with BCIs based on each simple approach (P300 and M-VEP). Offline data suggested that performance would be best in the combined paradigm. Online tests with adaptive BCIs confirmed that our combined approach is practical in an online BCI, and yielded better performance than the other two approaches (P<0.05) without annoying or overburdening the subject. The highest mean classification accuracy (96%) and practical bit rate (26.7bit/s) were obtained from the combined condition.
脑-机接口（BCIs）允许用户仅通过大脑活动进行交流。许多 BCI 依赖于当目标刺激闪烁时引发的 P300 和其他事件相关电位（ERPs）。尽管已经有大量研究探索了改进 P300 BCI 的方法，但令人惊讶的是，很少有工作关注改变视觉刺激以引出更可识别的 ERP 的新方法。在本文中，我们介绍了一种基于 P300 电位和运动起始视觉诱发电位（M-VEPs）的“组合”BCI，并将其与基于每个简单方法（P300 和 M-VEP）的 BCI 进行了比较。离线数据表明，组合范式的性能最佳。在线自适应 BCI 测试证实，我们的组合方法在在线 BCI 中是实用的，并且比其他两种方法（P<0.05）表现更好，而不会让受试者感到烦恼或负担过重。从组合条件中获得了最高的平均分类准确率（96%）和实用比特率（26.7bit/s）。
