Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.794693
Title: Optimising the quantitative analysis in functional pet brain imaging
Author: Alzamil, Yasser
ISNI:       0000 0004 8500 6160
Awarding Body: Cardiff University
Current Institution: Cardiff University
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Patlak analysis techniques based on linear regression are often applied to positron emission tomography (PET) images to estimate a number of physiological parameters. The Patlak equation forms the basis for most extension works regarding graphical analysis of many tracers in quantitative PET measurements. Patlak analysis is primarily used to obtain the rate constant Ki, which represents the tracer transfer rate from plasma to the targeted tissue. One of the most common issues associated with Patlak analysis is the introduction of statistical noise, adopted originally from the images, that affects the slope of the graphical plot, leading to bias, and causes errors in the calculation of the rate constant Ki i. In this thesis, several statistical and noise reduction methods for 2 and 3 dimensional data are proposed and applied to simulated 18F-FDOPA brain images generated from a PET imaging simulator. The methods were applied to investigate whether their utilisation could reduce the bias and error caused by noisy images and improve the accuracy of quantitative measurements. Then, validation step extended to 18F-FDOPA PET images obtained from a clinical trial for Parkinson's disease. The minimum averaged SE, SSE and the highest averaged reduction of noisy Ki values were found with the feasible generalised least squares (FGLS) model. Battle-Lemarie wavelet (BLW) showed significant change in data for the 3D PET images. Savitzky-Golay filtering (SGF) demonstrated significant change for most of the noise levels applied to 2D data. In clinical 18F-FDOPA images, the mean and standard deviation of standard error (SE) and sum-squared error (SSE) were significantly reduced in both baseline and after therapy groups. This work has the potential to be extended to other graphical analysis in quantitative PET data measurements.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.794693  DOI: Not available
Share: