Identification of Schizophrenia Using EEG Alpha Band Power During Hyperventilation and Post-hyperventilation

Abstract

The objective of the study is to analyze electroencephalograms (EEGs) of patients with schizophrenia using power spectral density. The proposed method measures various absolute powers that are related to the amount of information contained in the frequency components. 57 schizophrenia subjects and 24 normal subjects were included the study. EEG recordings were obtained under resting, hyperventilation and post-hyperventilation with eyes closed conditions. The absolute total power for 10 s epochs of EEGs was calculated using Welch’s method. The delta, theta, alpha, and beta bands were extracted from EEGs using a finite impulse response band pass filter and the power was determined. The differences in the absolute powers of the beta band and alpha bands were significant between the two groups (p < 0.001) during rest. The maximum power changes occurred in the delta and alpha bands when the brain condition changed from rest to hyperventilation and post-hyperventilation and this difference is more for normal subjects compared to the schizophrenia group. The subject groups were classified using a support vector machine classifier, yielding a high classification accuracy of 83.33% with 87.2% sensitivity and 82% specificity when alpha power was used as the input. These results suggest that schizophrenia subject can be identified using absolute alpha band power.