论文部分内容阅读
We use mutual information to characterize nonlinear correlations in the multi-channel scalp EEG of humans during the initial sleep cycle (52-90 min).Data from each channel are embedded and the mutual information of the embedded data is calculated using a new,locally adaptive partitioning method.We calculate the average mutual information of all 171 channel pairs in 19-ehannei scalp recordings during wakefulness,slow-wave sleep and REM sleep.