EEG Wavelet Spectral Analysis During a Working Memory Tasks in Stroke-Related Mild Cognitive Impairment Patients

Abstract

The aim of this study was to analyse the electroencephalography (EEG) background activity of 10 stroke-related patients with mild cognitive impairment (MCI) using spectral entropy (SpecEn) and spectral analysis. These spectral features were used to test the hypothesis that the EEG dominant frequencies slowdown in MCI in comparison with 10 age-match control subjects. Nineteen channels were recorded during working memory and were grouped into 5 recording regions corresponding to scalp areas of the cerebral cortex. EEG artifacts were removed using wavelet analysis (WT). The SpecEn analysis of the EEG data suggested a broad and flat spectrum in the normal EEG. The relative powers (RP) in delta (dRP), theta (?RP), alpha (aRP), beta (ßRP), and gamma (?RP) were calculated. SpecEn was significantly lower in stroke-related MCI patients at parietal, occipital and central regions (p-value < 0.05, Student’s t-test). Moreover, the other significant differences can be observed in increasing the dRP, ?RP and ?RP and decreasing the aRP and ßRP of the stroke-related MCI group in all regions (p-value < 0.05, Student’s t-test). It can be concluded that the SpecEn and spectral analysis are useful tool to inspect the slowing in the EEG signals in post-stroke MCI patients’ and the healthy controls’ EEG.