An Adaptive Virtual Keyboard with Visual and Auditory Feedback based the Detection of Steady-State Visual Evoked Potentials

Brain-Computer Interface (BCI) allows the direct communication between a brain and a computer through the translation of brain activity into commands. Non-invasive BCI using electroencephalography (EEG) recordings suffer from multiple drawbacks that prevent them from being deployed where a reliable performance is expected. BCIs based on the detection of visual stimuli require well designed user interfaces that take into account the constraints imposed by the visual stimuli characteristics and confusions that can occur between them. Efficient signal processing techniques are not enough for new and experienced users, the interface must be user-friendly. In this paper, we propose a new virtual keyboard with visual and auditory feedback based on the detection of steady state visual evoked potentials (SSVEP). We propose a tree-based menu layout where the position of the buttons is based on the distribution of the letters in the English letters, and where the visual stimuli allow to identify where the next desired command is placed in the interface. Furthermore, we propose an adaptive approach for determining the threshold to identify the commands, using the performance of the virtual keyboard. The evaluation supports the conclusion that the order of the commands has an impact on the time needed to select these commands.