Wavelet-Welch methodology for analysis of EEG signals of schizophrenia patients

Abstract

The goal of this study was to analyze whether schizophrenia patients may exhibit distinctive sub-band EEG features from control subjects compared both at rest and during auditory stimulation periods. EEG signals of thirty schizophrenic patients and age-gender matched healthy subjects were recorded from F3 left frontal region and analyzed using wavelet decomposition and Welch power spectral density (PSD) methods. Results show that PSD of all EEG sub-bands in patients was higher than that of control subjects during all periods of the experiment. Moreover, auditory stimuli evoked a significant decrease of Welch PSD of beta activity of EEG data in both groups. However, no significant change was found between stimulation periods in schizophrenia group. This methodology can be used to analyze EEG signals of schizophrenia patients to reach discriminative features between patients and controls.