ABSTRACT
Cerebral hemorrhage may cause cognitive dysfunction. Electroencephalogram (EEG) is a noninvasive diagnostic tool for assessment of cerebral function. A total of 174 patients (including cognitively impaired and cognitively normal) with cerebral hemorrhage, and 120 healthy persons (CN), were recruited between August 2008 and July 2012 at the Department of Neurology. EEG was used to analyze cerebral function of patients and normal persons. Correlation, clustering and concordance analyses were performed to analyze the relationship between EEG power and Montreal Cognitive Assessment (MoCA) scores. Cognitively impaired patients had a significantly decreased EEG beta power (0.793 ± 0.176 µV(2)) compared with cognitively normal patients (1.589 ± 0.205 µV(2), P < .01) or healthy persons (1.651 ± 0.185 µV(2), P < .01). Significantly negative correlations between beta power and hemorrhage region, size, amount, and patients' age were apparent (r = -0.91888, -0.78569, -0.84961, and -0.80365, respectively, all Ps < .001). There was good concordance between the K-means clustering algorithm calculating beta power and MoCA scoring (κ = .904, P < .001). In conclusion, the analysis method of EEG (beta power) abnormalities holds considerable promise to assess cognitive impairment after cerebral hemorrhage. Cognitive impairment was negatively correlated to hemorrhage region, size, amount, and age.
