&&ReWrAp:HEADERFOOTER:0:ReWrAp&&

Date of Award

6-1981

Degree Type

Thesis

Degree Name

Master of Engineering (ME)

Department

Electrical and Computer Engineering

Supervisor

S.K. Sarna

Abstract

Various tests had been conducted in order to examine the stationarity and the normality characteristics of electrical control signals of the digestive tract. They were done because the conventional frequency analysis, which is used extensively for the investigation of biological signals, usually assumes the signal to be stationary and normally distributed. The validity of this assumption should then be examined before any further analysis is applied. The tests are conducted by proposing a null hypothesis that the signal under investigation is stationary and normally distributed. It was found that the percentage of rejection of the hypothesis increases towards the colonic end of the tract.

Since conventional power spectral analysis does not provide any phase information on non-stationary signal, the bispectral analysis, which is the Fourier transform of the third moment; was used in order to examine many of the still-unknown characteristics of the gastrointestinal signal. The analysis mainly searches for any phase-locking between frequency components and hence identifies the generators of the signal. Two seperate analyses had been done : one was for a single channel and the other was for a double channel of signals. It was found that the same group of generators for the electrical signals on the upper part of the tract is present most of the time. But short-lived and locally based oscillators dominate the functions in the colon. From the cross-bispectral analysis, it was found that the generators in the stomach and the duodenum usually exert driving force to the distal site but bidirectionally in the jejunum. In the colon, only independent frequency components were found to be phased-locked occasionally.

In conclusion, the analyses carried out in this study provide some alternate means to investigate many of the still largely unknown signals.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS