Time-frequency methods for detecting spike activity of stomach

It has been hypothesised by many researchers that the spike activity signal s of the stomach are responsible for triggering peristaltic contractions. S ince most gastric motility disorders include an abnormality in the contract ion pattern, it is very important to access this information non-invasively . The aim in this study is to use abdominal electrogastrogram (EGG) signals to detect the spike activity signals generated by the serosa of the stomac h, and hence provide clinicians with a better method to monitor the motilit y state of the stomach. Through second and third-order spectral estimations performed on the serosal data obtained from canine experiments, it was con cluded that the spike activity in serosal signals occupies a frequency rang e of 50-80 cycles per minute. An increase in this frequency range during st rong antral contractions was observed both in the serosal and cutaneous pow er spectra. By using the 'continuous wavelet transform' with respect to a m odified Morlet wavelet, the spike activity signals generated from the seros a of the stomach can be detected and quantified in time from the cutaneous EGG records. During phase III contraction episodes, a detection accuracy of up to 96% from the cutaneous EGG recordings was calculated based on the sc ored serosal spike activities simultaneously recorded.