Application of physics and mathematics in clinical nuclear medicine

Ling, Lin (2013) Application of physics and mathematics in clinical nuclear medicine. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (12MB)

Abstract

This PhD research project consisted of two parts. In the first part a new pharmacokinetic model was introduced to improve the accuracy of kidney function estimation based on a sampling schedule of 2 and 4 blood sample measurements. Previous models such SET1 and SET2 have been shown to be unreliable with respect to the choice of sampling schedule and in some cases provide physiologically impossible values of the glomerular filtration rate (GFR). The new model called Tk-GV uses a Tikhonov regularized gamma variate function to fit the plasma clearance data. Based on a group of 46 patients, a comparison of four pharmacokinetics models (SET1, SET2, OLS-GV, and Tk-GV) revealed that the Tk-GV model was the most robust with respect to sample size and sampling schedule and provided no physiologically impossible values of G FR. When compared to constant infusion results in the literatures, the Tk-GV model was shown to eliminate the proportional overestimation of G F R produced by the SET1 model given by Chantler's correction and the constant overestimation produced by the SET2 model. -- Application of the Tk-GV model for estimating a patient's GF R requires the collection of 4 blood samples. In cases where it is not possible to collect four blood samples, two blood samples are collected and the patient's GFR is calculated using the SET1 model. A criterion called the renal sufficient index (RSI) was developed by others using a group of children to correct the overestimation of G F R as obtained from the SET1 model. In this work the RSI was applied to a mixed group of 26 patients (adults, children, males, and females) and shown, based on additional medical information, to be able to accurately distinguish between patients with normal and abnormal G FR. -- Bone mineral density (BMD) is important for evaluating bone health, especially for elderly people such as post-menopause women and men with prostate disease. However, analyzing a patient's BMD change is difficult because of naturally occurring short-term and long-term fluctuations in a patient's BMD, and also the short-term and long-term errors provided by the equipment used to measure BMD. The World Health Organization has provided a least significant change (LSC) criterion for distinguishing between normal and abnormal changes in a patient's BMD value. Unfortunately the LSC criterion only accounts for short-term machine error. In the second part of this thesis a new criterion called the total detectable difference (TDD) is introduced for analyzing changes in BMD in consideration of naturally occurring changes in the patient's BMD, and instrumentation error occurring on both short and long time scales. Based on the analysis of a group of 8,800 patients, a TDD value of approximately 0.045 g/cm² is obtained for the hip and 0.060 g/cm² for the lumbar spine. Based on the study of 9,379 patients, it is shown that the LSC criterion has a potential of over-diagnozing BMD change by approximately 30% compared to the TDD criterion.

Actions (login required)

View Item