Single Channel EEG Based Score Generation to Monitor the Severity and Progression of Mild Cognitive Impairment

摘要: Mild Cognitive Impairment (MCI) is a preliminary stage of Dementia. MCI determined by behavioral screening measures such as Montreal Assessment (MoCA) and Mini-Mental Status Examination (MMSE). Therefore, monitoring the progression predicting MoCA scores from objective physiological like EEG crucial it will not only help to improve mental healthcare aging population but also reduce costs. In this study, we demonstrate single channel based score generation method, which cost-effective suitable for continuous patient in longitudinal study. We collected scalp data while subjects were stimulated with five auditory speech signals. extracted 590 features Event-Related Brain Potentials (ERPs), included time spectral domain characteristics response. The top 11 features, ranked mutual information, used building regression models generate subjects. Robustness our model was tested using R-squared value, mean square error (MSE), residual's quantile plot, cook's distance. analysis shows R-squared=0.78 MSE=1.63, residual suggests that acceptable terms leverage, Cook's outcomes indicate single-channel can be estimate cognitive automatically severity detection monitoring, us efficaciously assess health status elderly people prognosis rehabilitation age-related impairments.