Real-time interfaces between the brain and a computer (brain-computer interface, BCI) have been used to restore motor functions lost through injury or disease. For example, progressive neurological diseases such as amyotrophic lateral sclerosis (ALS) can lead to severe or total loss of voluntary muscular control due to degeneration of central and peripheral motor neurons. As a result patients are severely or completely paralysed ­ a condition which is referred to as being 'locked-in'. These patients require alternative non-muscular methods for communication and control. Brain-computer interfaces measure specific features of the electrical activity of the brain and translate them into device commands. These commands do not depend on muscular control and can be used to operate an application. Thus, a BCI can provide communication and control for those who are locked-in. Electrical signals from the brain used for BCI operation can be recorded non-invasively from the scalp or invasively from the cortical surface. These signals include slow cortical potentials (1), P300 evoked potentials (2), and sensorimotor rhythms (3) recorded from the scalp. Some BCIs require self-regulation of the specific brain signal. Then patients are provided with online feedback of their EEG and knowledge of correct responses. In subsequent trials patients have to move a graphic signal (cursor) on a monitor toward targets located at the top or bottom of the screen. We trained severely or totally paralysed patients to regulate their EEG and to use this ability for communication with the aids of a Language Support Program. All patients were trained at home and some communicated extensive messages (4). Brain-computer interfaces for severely paralysed patients have been mainly used for verbal communication. A BCI controlled browser to surf the internet is also available. Furthermore, brain-computer interfaces were used to restore grasping via functional electric stimulation (5). The usefulness of BCIs to maintain or reinstall communication or other lost motor function in locked-in patients has been frequently demonstrated. However, signal analysis has to be improved, training protocols simplified and a better knowledge of the psychosocial factors interacting with BCI use is crucial to make BCI technology ready for routine clinical application.



(1) Kübler A et al. Exp Brain Res 1999;124:223-232.
(2) Donchin E et al. IEEE Trans Neural Syst Rehab Eng 2000;8(2):174-9.
(3) Wolpaw JR et al. IEEE Trans Neural Syst Rehab Eng 2003;11(2):204-7.
(4) Neumann N et al. J Neurol Neurosurg Psychiatry 2003;74(8):1117-21.
(5) Pfurtscheller G et al. Neurosci Lett 2003;351(1):33-6.