Automatic user customized brain switch

Abstract

Designing a self-paced brain computer interface (BCI) that works reliably across human subjects has been a challenge. Of particular interest are simple BCIs that enable detection of an Intentional Control (IC) state against a background of No Control (NC) state. Such BCIs are known as brain switches or BCI switches. One of the possible methods to build a BCI switch is based on the consistent increase in the alpha component of the EEG spectrum when subjects close their eyes. The present work proposes a simple approach to achieve automatic user customization with just one minute of calibration recording. This design is evaluated based on thirty trials on each of the seven healthy subjects. The results validate the robustness of the proposed IC detection approach to variations across sessions and subjects. The average time required to activate the switch is typically two to three seconds after eye closure. These results suggest that any user with the ability to modulate their alpha rhythm by eye closure can effectively utilize the proposed BCI switch as an assistive technology to gain increased control over their environment.