Return to search

On the Accuracy of Wireless Capsule Endoscope RF and Visual Localization

Wireless capsule endoscope (WCE) is becoming one of the most patient-friendly inspection device which provides visual investigation of entire gastrointestinal (GI) tract, while the other traditional (wired) endoscopic devices are usually designed for colon inspection. Locating abnormalities tract such as tumors, polyps and bleedings with wire-connected endoscope in GI tract is simple as long as we could measure the length of the wires inside human body. When WCE is applied, however, this becomes a critical challenge of examination since there is no wires connected to WCE while physicians need to find the exact locations of WCE to identify the position of abnormalities. To locate the WCE accurately, methods have come up in last decade including time of arrival (TOA) based methods, received signal strength (RSS) based methods, phase difference of arrival (PDOA) based methods, electromagnetic methods and video-based tracking methods, etc.. In this thesis, time of arrival (TOA), phase difference of arrival (PDOA) and video based localization methods are proposed and their performance are analyzed. We first propose an novel video-based tracking technique based on maximum mutual information. With this technique, we can tell the displacement and rotation between consecutive frames. Then in TOA chapter, the Cramer-Rao lower bound (CRLB) of TOA ranging inside homogeneous tissue is calculated first then three TOA ranging methods are proposed and compared with the CRLB which is used as the performance guideline. After that, PDOA based ranging technique is applied exploiting phase difference of two signals. Since the phase difference is taken into consideration, the ranging ambiguity is eliminated. We also evaluate the performance of the proposed PDOA ranging method. Finally, these ranging methods are evaluated in non-homogeneous tissues, the results of which are also compared to that in homogeneous tissue to analyze the impact of non-homogeneity.

Identiferoai:union.ndltd.org:wpi.edu/oai:digitalcommons.wpi.edu:etd-theses-1516
Date29 April 2015
CreatorsZhou, Mingda
ContributorsKaveh Pahlavan, Advisor, Allen H. Levesque, Committee Member, Xinming Huang, Committee Member, Emmanuel O. Agu, Committee Member, Steve Howard, Committee Member, Guanqun Bao, Committee Member
PublisherDigital WPI
Source SetsWorcester Polytechnic Institute
Detected LanguageEnglish
Typetext
Formatapplication/pdf
SourceMasters Theses (All Theses, All Years)

Page generated in 0.0021 seconds