Entropy-Based EEG Markers for Gender Identification of Vascular Dementia Patients

摘要

The efforts in this study were being made to understand how gender differences contribute to identifying the early signs of cognitive impairment that were lead to severe dementia.The electroencephalogram(EEG)of 5 patients with vascular dementia(VD),15 stroke-related mild cognitive impairment(S_MCI)patients,and 15 healthy control(HC)subjects during a working memory task(WMT).To do so,Savitzky-Golay(SG)filter was applied in the first stage as a preprocessing method for the seek of EEG signals smoothing and artifacts removal.In the second stage,four different entropies were extracted from the denoised EEG signals to test the hypothesis of reducing the complexity in both VD and SJvlCI in comparison with HC these are spectral entropy(SpecEri),fuzzy entropy(FuzEri),tsallis entropy(TsEri)and improved permutation entropy iimpe).In the next step,statistical analysis has been conducted by using multivariate analysis of variance(MANOVA)to assess the gender differences over the brain regions for S_MCI and dementia patients compared with HC subjects.Results show that the entropy-based EEG analyses may provide a simple method with a cost-effective way to identify and quantify the severity of dementia patients.In conclusion,increased dementia severity was associated with decreased entropy-based features complexity.Thus,EEG could be the key to report interesting information for differentiation the EEG background activity in female and male of patients with VD and S_MCI to help medical doctors to identify gender differences effects on dementia survivors.