Using EEG/MEG Data of Cognitive Processes in Brain-Computer Interfaces

摘要

Brain-computer interfaces (BCIs) aim at providing a non-muscular channel for sendingcommands to the external world using electroencephalographic (EEG) and, more recently, magne-toencephalographic (MEG) measurements of the brain function. Most of the current implementa-tions of BCIs rely on EEG/MEG data of motor activities as such neural processes are well charac-terized, while the use of data related to cognitive activities has been neglected due to its intrinsiccomplexity. However, cognitive data usually has larger amplitude, lasts longer and, in some cases,cognitive brain signals are easier to control at will than motor signals. This paper briefly reviews theuse of EEG/MEG data of cognitive processes in the implementation of BCIs. Specifically, this pa-per reviews some of the neuromechanisms, signal features, and processing methods involved. Thispaper also refers to some of the author's work in the area of detection and classification of cognitivesignals for BCIs using variability enhancement, parametric modeling, and spatial filtering, as wellas recent developments in BCI performance evaluation.