ABSTRACT
The basic theory of Higher Order Spectral Analysis and the most generally used Bispectrum are introduced in the paper. By certain experiments of EEG signal acquisition and bispectrum analysis, it is showed that the Higher Order Spectrum has an advantage over power spectrum, which is based on Second Order Statistics, in processing non-linear signal and restraining Gauss noise signal.
Subject(s)
Electroencephalography , Methods , Signal Processing, Computer-AssistedABSTRACT
Many researches have proved that auditory evoked potentials (AEP) can indicate the depth of anesthesia in the clinical practice. This paper introduces two methods of extracting AEP: the moving time average (MTA) model and the autoregressive model with exogenous input (ARX). In our study we used both the MTA model and the ARX model to extract AEP and the results show that using the ARX model can get the results significantly faster than using the MTA model, and the ARX model is more suitable for the real-time monitoring of the anesthetic depth.