PolyU IR
 

PolyU Institutional Repository >
Electronic and Information Engineering >
EIE Theses >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/3856

Title: Advances in functional imaging with emission computer tomography
Authors: Lau, Chi-hoi
Subjects: Diagnostic imaging
Tomography, Emission
Hong Kong Polytechnic University -- Dissertations
Issue Date: 1999
Publisher: The Hong Kong Polytechnic University
Abstract: With the advent of technology for emission computer tomography, functional imaging technique, using either Positron Emission Tomography (PET) or Single Photon Emission Computer Tomography (SPECT), provides a powerful tool to yield invaluable information of the physiological processes under study. Based on this technique, quantitative portrayal of both structure and function of these processes are obtained. It has created opportunities for researchers to examine increasingly complex biomedical systems, thereby leading the way to a deeper understanding of fundamental complexities of life. It has also greatly improved the diagnosis and treatment of human diseases. In spite of the wide applicability in both scientific researches and clinical diagnostic procedures, functional studies are still of limited uses due to a number of factors. For example, the cost of PET scanners is much expensive than the other imaging modalities, which make them not commonly available in most hospitals. Although effort has been made in studying the use of less expensive SPECT scanner to perform the similar dynamic studies, the incompleteness of the projection data introduces artifacts to the reconstructed images which will affect the accuracy of subsequent parameter estimation. Furthermore, little attention has been paid to systematically investigate the scanning duration which usually determined based on clinician's intuition and experience. The slow kinetics of typical SPECT tracers may require unrealistically long total acquisition times to obtain reliable estimation of the slower rate constants. Finally, dynamic PET studies require the measurement of tracer concentration in blood plasma which is invasive, time-consuming and tedious. The blood measurement process also exposes the clinical personnel to the danger of fatal blood infection and radiation. In this thesis, three different techniques were presented in order to solve the aforementioned limitations. Our primary focus is to develop computationally efficient techniques to minimize the inconvenience of performing functional imaging using emission computed tomography, by means of either PET or SPECT.
In the first part of the thesis, we investigated the possibility of using rotating detector systems to perform functional studies, rather than full ring detector systems. The rotating detector systems are usually of much lower cost than the full ring detector systems. We study the problem of parameter estimation using measurements recorded by rotating detector systems and propose an approach to improve the accuracy of parameters estimated using the above detector systems. The method involves interpolation across projections so as to provide an improved estimate of the projections, closer to those which would be obtained for a stationary detector system. The interpolated projections are then reconstructed using the conventional filtered back-projection (FBP) algorithm and the kinetic parameters are estimated using a weighted least squares cost function based on the integral of activity. The proposed method is based on several previously validated techniques which, in combination, provide a simple and computationally efficient solution. The second part of the thesis is to systematically investigate the reliability of parameter estimation as a function of total acquisition time, with Thalium-201 (T1-201) dynamic SPECT as our typical example. We also suggest and evaluate a clinically practical alternative to prolonged continuous dynamic acquisitions. In addition, the minimum number of frames and their duration are investigated using the optimum sampling schedule (OSS) technique. While this study concentrates on applying the methodology to T1-201 kinetics, the methodology developed here is also applicable to other dynamic SPECT studies. In the final part of the thesis, we propose an approach to estimate the parameters of a pre-assumed tracer kinetic model when performing dynamic studies, eliminating the requirement of measuring the tracer concentration in blood sampling. The approach involves a denoising step for the projection data. In our case, we use a wavelet denoising approach, given its advantage of preserving the structural information of the image when filtering the noise. The denoised projections are then reconstructed using the FBP. From the reconstructed dynamic images, tissue time activity curves (TTACs) are extracted from regions of interest (ROIs) and are re-sampled using linear interpolation. Based on these TTACs, an eigen-vector based blind deconvolution technique is applied to estimate the kinetic parameters. On the whole, the aim of these studies is to develop computationally efficient algorithms to reduce the inconvenience associated with functional studies using emission computed tomography. We hope that such studies can be more commonly and extensively used in both scientific researches and clinical environments.
Degree: M.Phil., Dept. of Electronic and Information Engineering, The Hong Kong Polytechnic University, 1999
Description: xv, 113 leaves : ill. ; 30 cm
PolyU Library Call No.: [THS] LG51 .H577M EIE 1999 Lau
Rights: All rights reserved.
Type: Thesis
URI: http://hdl.handle.net/10397/3856
Appears in Collections:EIE Theses
PolyU Electronic Theses

Files in This Item:

File Description SizeFormat
b14799364_ir.pdfFor All Users (Non-printable) 4.48 MBAdobe PDFView/Open
b14799364_link.htmFor PolyU Users 162 BHTMLView/Open



Facebook Facebook del.icio.us del.icio.us LinkedIn LinkedIn


All items in the PolyU Institutional Repository are protected by copyright, with all rights reserved, unless otherwise indicated.
No item in the PolyU IR may be reproduced for commercial or resale purposes.

 

© Pao Yue-kong Library, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong
Powered by DSpace (Version 1.5.2)  © MIT and HP
Feedback | Privacy Policy Statement | Copyright & Restrictions - Feedback