Channel selection and feature extraction for cognitive state estimation with the variation of brain signal

Cognitive state estimation with the variation of physiological signal has drawn extensive attention from disciplines such as psychology, cognitive science and engineering. In this paper, we present a cognitive state classification system to assess the subject's mental states based on EEG measurements. The cognitive state estimator is utilized in the context of an augmented cognition system that aims to enhance the cognitive performance of a human user through computer-mediated assistance based on EEG. This paper focuses on the channel selection of the BIOPAC automated EEG analysis and feature extraction based on spectral analysis. Different frequency components (i) real value; (ii) imaginary value; (iii) magnitude; (iv) phase angle and (v) power spectral density of the EEG data samples in different mental conditions are selected for the estimation of cognitive states. In this approach seven types of cognitive states such as — relax, mental task, memory related task, motor action, pleasant, fear, and enjoying music are selected for feature extraction in the three channels — EEG, Alpha, and Alpha RMS of BIOPAC EEG data acquisition system. After feature extraction the channel efficacy are evaluated by support vector machine (SVM) which is based on the classification rate in different cognitive states. From experimental results and classification accuracy, it can be determined that alpha channel can be selected for cognitive state estimation. The overall accuracy for alpha channel shows much improved result for power spectral density and the classification rate is 69.17% whereas for EEG and alpha RMS channel it is found 47.22% and 32.21% respectively.