A Novel Task-Oriented Optimal Design for P300-based Brain–computer Interfaces

Zongtan Zhou,Erwei Yin,Yang Liu,Jun Jiang,Dewen Hu
DOI: https://doi.org/10.1088/1741-2560/11/5/056003
IF: 5.043
2014-01-01
Journal of Neural Engineering
Abstract:Objective. The number of items of a P300-based brain-computer interface (BCI) should be adjustable in accordance with the requirements of the specific tasks. To address this issue, we propose a novel task-oriented optimal approach aimed at increasing the performance of general P300 BCIs with different numbers of items. Approach. First, we proposed a stimulus presentation with variable dimensions (VD) paradigm as a generalization of the conventional single-character (SC) and row-column (RC) stimulus paradigms. Furthermore, an embedding design approach was employed for any given number of items. Finally, based on the score-P model of each subject, the VD flash pattern was selected by a linear interpolation approach for a certain task. Main results. The results indicate that the optimal BCI design consistently outperforms the conventional approaches, i.e., the SC and RC paradigms. Specifically, there is significant improvement in the practical information transfer rate for a large number of items. Significance. The results suggest that the proposed optimal approach would provide useful guidance in the practical design of general P300-based BCIs.
What problem does this paper attempt to address?