Solving reality problems by using mutual information analysis

Abstract

Cross-mutual information (CMI) can calculate to time series for thousands of sampled points from corticocortical connection among different functional states of brain in Alzheimer's disease (AD) patients. The aim of this study was to use mutual information analysis in the multichannel EEG to predict the probability of AD disease. Considering the correlation between AD disease and ageing effect, the participants were 9 AD patients and 45 normal cases involving teenagers, young people and elders. This data revealed that both right frontal and temporo-parietal are differences between normal and AD participants. Besides, this study found the theta band is the main frequency to separate AD patients from all participants. Furthermore, this study suggested a higher distinguishable method by mutual information to predict the possibility AD patients. © 2014 Chia-Ju Liu et al.