Give Me a Sign: Studies on the Decodability of Hand Gestures Using Activity of the Sensorimotor Cortex as a Potential Control Signal for Implanted Brain Computer Interfaces

© The Author(s) 2014 C. Guger et al. (eds.), Brain-Computer Interface Research, SpringerBriefs in Electrical and Computer Engineering, DOI 10.1007/978-3-319-09979-8_2 The major driving force behind the development of brain computer interfaces (BCI) has been the desire to re-establish communication for severely paralyzed or even locked-in patients. As a consequence, different strategies have been developed to provide a direct link between a still-functional brain and the outside world, bypassing the non-functional muscle system (Wolpaw et al. 2002). The first BCIs used the P300 evoked potential (Farwell and Donchin 1988) and slow cortical potentials (Birbaumer et al. 1999), as they can be measured by electroencephalography (EEG). In the ideal case, a BCI would enable its user, previously incapable of any communication, to participate in a conversation at the same speed and with the same expressiveness as a non-paralyzed person would. However, the use of electroencephalography (EEG) as the primary recording method has to date limited the potential to decode brain activity, due to its low spatial resolution and signalto-noise ratio. EEG can only detect prominent changes in brain activity and often has to integrate information over time in order to detect a certain activity pattern. These limitations reduce the speed and flexibility of any communication based on EEG BCI.