Asynchronous P300-Based Brain--Computer Interfaces: A Computational Approach With Statistical Models

摘要: Asynchronous control is an important issue for brain--computer interfaces (BCIs) working in real-life settings, where the machine should determine from brain signals not only desired command but also when user wants to input it. In this paper, we propose a novel computational approach robust asynchronous using electroencephalogram (EEG) and P300-based oddball paradigm. approach, first address mathematical modeling of target P300, nontarget noncontrol signals, by Gaussian distribution models support vector margin space. Furthermore, derive method compute likelihood state time window EEG. Finally, devise recursive algorithm detect states ongoing EEG online application. We conducted experiments with four subjects study both BCI's receiver operating characteristics its performance actual tests. The results show that BCI able achieve averaged information transfer rate approximately 20 b/min at low false positive (one event per minute).