Classification of resting, anticipation and movement states in self-initiated arm movements for EEG brain computer interfaces.
Accurately identifying cognitive activity from multichannel EEG data continues to be a challenging task for cognition researchers. Although brain impairments such as epilepsy and ADHD tend to display relatively easy to identify EEG data features, delineating clear patterns of normal cognitive activity within the healthy brain has not yet been satisfactorily achieved. In the current study, the EEG features associated with distraction have been examined. Continuous EEG has been recorded from eight participants as they performed normal and auditory-distraction virtual driving. All pole analysis using the autoregressive model of the EEG data in both conditions has been then carried out, producing interesting and anatomically feasible patterns in temporal lobe regions. Event related synchronization and desynchronization patterns have also been observed at specific brain locations within the alpha, beta and gamma band frequencies. These results clearly distinguish audio-distracted driving from normal driving, suggesting that the techniques used here may provide a new analysis method for brain computer interface research. Keywords-Modeling EEG; cognition; ERD/ERS;distracted driving.