Multiscale Entropy Analysis of EEG Signals

Electroencephalogram (EEG) contains a lot of pathological information, which is the most effective clinical tool for detecting epileptic seizure. In this paper, multiscale entropy is used to analyze and compare normal EEG signals and epileptic EEG signals in detail to study whether multiscale entropy can effectively distinguish EEG signals in different states. The results show that multiscale entropy can correctly distinguish between normal EEG signals and epileptic EEG signals, and the entropy value can effectively reflect the seizure interval and seizure state, so as to predict seizures.