Year

2005

Degree Name

Doctor of Philosophy

Department

School of Physics - Faculty of Engineering

Abstract

This study is focused on the analysis, evaluation and calibration of low cost computer desktop scanners for possible use in film dosimetry. The study also includes the development a software package to use the digital output of such scanners for image processing. This software has matured commercially into a suite of programs named Radiation Oncology Dosimetry Management System (RODOMS) for use in radiotherapy quality assurance protocols. The input to the software is a bitmap image created by the scanner. The scan signal greyscale value (GSV) is converted into an optical density value (OD) and a calibration is made to the equivalent radiation doses (DOSE). The software functions include: �� Scanner signal calibration - using a standard step wedge film to set and calibrate the dynamic range and scan signal linearity �� OD vs DOSE response curve calibration - using film strips to read the optical density for different dose exposures to establish the OD-DOSE conversion curve. �� Background level uncertainty control - analysis and subtraction of the base + fog value from the film base material and the film over response to radiation scatter. �� Processing noise smoothing - polynomial and mean smoothing, alternately used to reduce the noise caused by film artifacts. �� Beam quality dependence correction - using individual OD-DOSE calibration curves to correct the exposure for a combined beam modality field film dosimetry quality assurance. �� Clinical film dosimetry analysis - graphic user interface (GUI) designed program for radiation field quality assurance in advanced comparison of 3-D and IMRT dose distribution analysis. This includes the IMRT field with the planning computer curve overlaid for a comparative analysis. Dosimetry phantoms have been designed and used in the testing and evaluation of RODOMS. The software is currently in clinical use in several radiotherapy centers in Australia and Asian countries. The clinical results give on average a ±3% uncertainty level in most of the clinical cases compared with the ionization or TLD measurement results of up to ±5%. By using this software in a variety of clinical situations it is shown in this study that the traditional uncertainty levels in film dosimetry have been significantly reduced.

02Whole.pdf (5061 kB)

Share

COinS
 

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.