Degree Name

Doctor of Philosophy


Centre for Medical Radiation Physics


Microbeam radiation therapy (MRT) is an experimental radiosurgical technique under investigation as a treatment for presently incurable brain lesions such as gliomas. The treatment involves the irradiation of a macroscopic target volume by delivering a very high dose to many microscopic wide sub-volumes. A very high dose rate and very low beam divergence are prerequisites for MRT, which at present necessitates the use of synchrotron radiation. The complex structure of the radiation field combined with the very high dose rate pose a challenge for conventional dosimetry. To address the dosimetric inadequacies, a complete and standalone dosimetry system was developed for MRT as part of this work. It includes the silicon microstrip detector: a high spatial resolution single strip detector, data acquisition hardware capable of remotely biasing and reading out the detector, and data acquisition software to interface with the hardware and perform analysis on received data. The system is collectively known as X-tream.

This thesis provides an introduction to MRT, the production of synchrotron X-rays and a literature review of methods of dosimetry for MRT as reported in literature. The development of the X-tream system is discussed along with an overview of its operation and initial characterisation.

Charge-collection characterisation of the microstrip detector was performed in numerous ways. Alpha spectroscopy reveals that the application of the guard ring greatly reduces the effective sensitive volume of the detector when compared to the floating configuration, but reveals no spatial information. Spatially-resolved ion beam induced charge collection (IBICC) is performed using the ANSTO Heavy Ion Microprobe to obtain spatial information about the charge collection distribution under various electrical configurations. The application of the guard ring is seen to greatly confine charge collection to the immediate vicinity of the central strip, however significant charge collection is seen near the pad region, a phenomenon not observed during MRT measurements. As the detector was not pre-irradiated prior to the investigation, it is hypothesised the combination of the metallic pad, oxide layer and silicon acts as a MOS capacitor and induces an electric field bubble in the region surrounding the pad. Subsequent exposure to radiation is then hypothesised to result in a reduction of the carrier lifetime and decrease the sensitivity of the detector. Synchrotron X-ray beam induced charge collection studies are performed on the ESRF ID17 beamline, and agree with the tight charge confinement of the guarded detector as observed for IBICC. Charge collection is not seen in the pad region, but the notable difference is that the detector has been pre-irradiated.

Energy dependence studies are performed through characterisation with a clinical orthovoltage X-ray unit and compared to Monte Carlo simulation. Both experimental and simulated results show an over-response of the detector to low energy X-rays when compared to dose to water. A more rigorous treatment is applied experimentally with monochromatic synchrotron X-rays over the range of 40 to 80 keV, which too reveal an over-response at low energy, a finding backed by Monte Carlo simulation. Hypothetical detector designs are simulated to attempt to reduce energy dependence, which include a reduced epitaxial layer thickness, a mesa structure and the use of CVD diamond. The reduced epitaxial thickness makes a minor difference, whereas the mesa structure and use of CVD diamond more dramatically reduce the low energy over-response.

The system is characterised in a synchrotron X-ray MRT field where it is found the detector requires pre-irradiation of the order of 20 kGy prior to use. Connecting the guard ring is found to greatly decrease detector sensitivity due to much better confinement of the charge collection volume. The system is found to have excellent linearity with prescribed dose, however depth dose measurements deviate from ionisation chamber measurements and Monte Carlo simulation of both dose to detector and dose to water. This is attributable to the energy dependence of the detector to low energy photons. With the detector oriented such that the smallest sensitive volume dimension is in the dose gradient direction, the substrate is found to result in asymmetric microbeam dose profiles. This is due to the dose enhancement of the silicon substrate relative to the phantom material on the opposite side.

The capabilities of the X-stream system are demonstrated in synchrotron X-ray MRT fields. The over response dependence of the microstrip detector to low energy photons however necessitates improvements to its energy dependence relative to water. A move to mesa-based structures and use of materials such as CVD diamond are suggested. The modularity of the X-tream system ensures that such changes of design require only minor modification to the rest of the system.