University of Adelaide
An optical computed tomography scanner for three dimensional gel dosimetry of radiotherapy dose distributions
Abstract
dc:description.abstractClinical radiotherapy treatments using linear accelerator (linac) generated megavoltage xray beams are planned using computer models that calculate patient specific three dimensional (3D) radiation dose distributions. Treatment planning system (TPS) calculated doses are evaluated by clinicians to ensure suitable dose coverage of targeted tumours and the avoidance of excessive doses to normal tissues. The accuracy of the TPS must be validated by measurement to ensure correct patient treatments. Traditional radiotherapy dosimeters do not measure dose entirely in 3D. They effectively ‘spot check’ accuracy at discrete points or planes, without the ability to fully visualise measured dose distributions in true 3D. True 3D dosimetry systems have been a subject of research for more than 3 decades. Gel dosimetry with optical computed tomography (CT) scanning using visible light wavelengths has been under investigation and development for much of this time. A lack of clinical uptake of the systems developed to date suggests that there are deficiencies or unappealing aspects, such as optical CT scanner maintenance and reliance upon optical expertise. Dosimetric accuracy of these systems also requires improvement, closer to accepted clinical dosimeters. In this work it was postulated that an optical CT scanner could be developed that is more efficient, practical and accurate than those demonstrated previously. This would address key aspects relating to clinical appeal. A specific application of stereotactic radiosurgery was targeted, where small dose distributions are delivered with high spatial accuracy to cranial tumours. Improvements in the practicality and efficiency of optical CT scanning were initially sought by elimination of the need for a refractive index (RI) matching fluid bath for scanning gel dosimeter samples. Optical simulations were used to investigate and identify suitable optical geometry that would enable fluid-less scanning for tomographic reconstruction. A prototype fluid-less optical CT scanner was constructed and the proof of concept was demonstrated using ferrous xylenol orange gel (FXG). The next phase of work was to develop the dosimetry system further, striving for dosimetric accuracy. The properties of the FXG dosimeter were studied by characterising the dose response, dose development, ion diffusion and thermochromism. The scanner was refined by addressing imaging artefacts, the addition of a reference detector and the development of a cuvette based dose calibration procedure. Standard procedures for gel manufacturing and handling, and for optical CT operation were developed to improve reproducibility of results. The system’s performance was assessed and its utility was demonstrated in the clinical application of linac radiosurgery and was also extended to brachytherapy 192Ir source verifications. Further improvements of system practicality and measurement quality were proposed by the concept of dual wavelength scanning. A second laser of a different wavelength could provide reference scan data instead of pre-irradiation scans. This would give a single scan procedure for greater practicality and could improve measurement quality by avoiding scan to scan artefacts. Dual wavelength feasibility was first investigated by revisiting optical simulations and the development of a design incorporating a second laser. Through the addition of the second wavelength further insights were gained into optical artefacts and image quality improvements were realised. An alternative calibration method using a reconstructed test pattern was compared to the cuvette method and resulted in improvements of calibration accuracy for dual wavelength scanning. The dual wavelength scanner dosimetry system was tested and benchmarked using a range of test irradiations, with comparison to reference dosimeters. This culminated in a true 3D dosimetry solution for radiotherapy dose verifications with accuracy of the order of 1 %, together with practical and efficient optical readout for clinical use.
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Ramm, Daniel Paul Norman
- Advisor dc:contributor.advisor
-
- Penfold, Scott
Subjects
dc:subject × 4Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- http://hdl.handle.net/2440/119799
- OAI identifier oai:identifier
- oai:digital.library.adelaide.edu.au:2440/119799