BCI Brain Computer Interface / BMI Brain Machine Interface

A BCI allows gaining information about the cognitive state of a human being without the necessity to produce any activity outside the central nervous system.

The first approach of using brain activity directly for communication comes from Vidal in the early 70s. First steps into the field of application have been realized by Birbaumer and Wolpaw in the 80s and 90s by building EEG-based support systems for patients suffering from amyotrophic lateral sclerosis (ALS).

The application of methods from statistical machine learning (Blankertz, Curio & Mueller 2002) for BCI had a deep impact on the accuracy of detecting patterns within the EEG data. Subsequently, it was possible to transfer the learning effort from the human being to the machine. Hence, it minimized the effort that is needed to develop the skills to control a BCI system.

When the focus is laid on the applicability of BCI while interacting in typical human-machine system environments the term Brain-Computer Interaction often is used instead of Brain-Computer Interface. User states hardly inferable from exogenous factors are estimated by the BCI, interpreted and fed back into the technical part of the system. The information about the user states can be handled as explicit (active BCI) or implicit (passive BCI) commands (Zander 2008).