I. Introduction
Most of the Brain-Computer Interfaces (BCIs) are currently based on electroencephalography (EEG) recordings as EEG amplifiers can be portable and wireless. However, despite the possibility to obtain a high spatial resolution with high density EEG (e.g., 128 channels), it requires a long preparation time, which can become an obstacle when dealing daily with BCI or when an application requires a high attentional load and the preparation time absorbs the focus of the participant, limiting his/her attention during signal recording. Although dry EEG solutions have been recently proposed, they don't provide a high spatial resolution due to the size of the sensors that are used. In clinical settings where the quality of the signal is a major concern and there is no constraint related to the portability, other brain imaging techniques can be considered, such as magnetoencephalography (MEG), which has, despite its current cost, advantages over EEG signals.