Improving the Performance of Brain-Computer Interface Using Multi-modal Neuroimaging

Non-invasive brain-computer interfaces (BCIs) allow users to control external devices by their intentions. Nevertheless, most current BCI systems rely on cues or tasks to which the subject has to react (i.e., synchronous BCIs). Such systems have limited applications in the real world. It is more desirable for the user to decide himself, when he likes to control a device. However, these so-called asynchronous BCI systems, that rely on electroencephalogram (EEG) measurements show the demand for higher accuracy and stability. Previously, hybrid BCI systems, relying on simultaneous EEG and near-infrared spectroscopy (NIRS) measurements, have been shown to increase the classification performance of (synchronous) motor imagery (MI) tasks. Here we present the first asynchronous hybrid BCI with encouraging results.