Time-frequency analysis of the electrogastrogram: evaluation of STFT, adaptive spectral analysis and exponential distribution

Abstract

The aim of this paper is to compare the performance of the short-time Fourier transform (STFT), adaptive spectral analysis and exponential distribution (ED) methods in estimating the changes of both amplitude and frequency of the gastric slow wave with time, and detecting gastric dysrhythmic events from the electrogastrogram (EGG). Results from both computer simulations and real applications show that the adaptive spectral analysis method is more accurate in the detection of gastric dysrhythmia with brief duration, whereas STFT method is more appropriate for the estimation of signal amplitude. The performance of the ED method appears to be in between the STFT method and adaptive spectral analysis method.