Feasibility analysis and adaptive thresholding for mobile applications controlled by EEG signals

Given the availability of EEG technology and existing studies, this paper discusses the feasibility of development of mobile applications controlled by brainwaves using a low-cost, non-invasive, headband type of device that collects two-channel EEG signals at frontal lobe. We have performed temporal, spectral and spatial analysis on EEG signals collected during game-playing and found particular trends of EEG signals at certain brain (mental) states for all subjects, and some variations of the trends among different subjects. The analysis results motivate us to design an adaptive thresholding mechanism to find user-specific thresholds for a classifier that controls mobile applications.