David Trocellier wins “best poster award” at the 9th International Graz Brain-Computer Interface conference.
In this paper, we evaluate on a large and open-source dataset different predictors of Brain computer interface (BCI) performance based on neurophysiological predictor. BCI allow to extract command based on electrical brain activity measured with electroencephalogram (EEG). However, individual performance in controlling a BCI varies significantly among participants. Neurophysiological predictors, extract from EEG signal during or before the task, can be good predictor of this variability. In this paper we aveluate validate six predictors proposed previously in the literature, only four of the six predictors has been validated on this database.