Detection of alertness states using electroencephalogram and cortical auditory evoked potential responses

In this paper, we focus on identifying the alertness state of subjects undergoing the cortical auditory evoked potential (CAEP) hearing test. A supervised classification approach is adopted, where subjects were advised to indicate their alertness states in specified time instances. Two sets of features are considered here to represent the recorded data. The first is based on the wavelet transform of the background EEG, while the second is obtained from the peaks of the CAEP responses. The rational behind using the second feature set is to evaluate the relationship between CAEP responses and alertness levels. Obtained results suggest that the CAEP-based features are very comparable, in terms of classification accuracy, to the well-established wavelet-based features of EEG signals (79% compared to 80%). The findings of this paper will contribute towards a better understanding of CAEP responses at the different alertness states.