Year

2019

Degree Name

Doctor of Philosophy

Department

School of Physics

Abstract

The integration of a linear accelerator with magnetic resonance imaging, or MRI-linac, provides a means to acquire gated images of the tumour and surrounding organs. This may enable more accurate radiation therapy treatment with the potential for motion-tracking using these images. Challenges of this kind of treatment include effects due to the magnetic and radiofrequency (RF) fields on the dose distribution. Magnetic fields in uence charged particles via the Lorentz force. The `Electron Return Effect' (ERE) and asymmetry of the point spread kernel are dosimetric effects due to a magnetic field that is transversely orientated with respect to the radiation beam. Narrower penumbral widths result from an inline magnetic field, as well as an increase in skin dose. Furthermore, for validation of this dose delivery, radiation detectors that can operate under magnetic and RF fields are needed. The effect that these fields have on the detector response, as well as the deposited dose, is investigated.

A QA concept named `MR dynamic dosimaging' is introduced with the potential to combine the MagicPlate-512 (M512) silicon array detector with MRI-compatible phantoms on a moving platform. MR visible/compatible detector systems are required for this to be achieved. Preliminary studies towards dosimaging were completed separately on a clinical MRI scanner and standard linac. A type of imaging hydrogel, named `gel-water' was investigated to enable the M512 to be MR visible, and at the same time replace solid water/water as a dosimetry phantom. It was concluded that gel-water is a feasible option to achieve these aims in an MRI-linac. The system with gel-water and the M512 had an agreement with solid water of ≤1.3% for various dosimetric characterisation measurements.

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.