Investigation into the dosimetric characteristics of MOSFETs for use for in vivo dosimetry during external beam radiotherapy.

Abstract

This thesis investigates the response to ionising radiation, of p-type Metal Oxide Semiconductor Field Effect Transistors (MOSFETs) (REM Oxford (UK)) and a reader system developed by the Centre for Medical Radiation Physics, The University of Wollongong, to determine their feasibility for measurements of dose during radiotherapy treatment (in vivo dosimetry (IVD)). Two types of MOSFET probes were used -"single sensitivity", for measuring low doses, and "dual sensitivity", to measure both high and lose doses. Sensitivity, linearity of response with dose, and response changes with accumulated dose and direction of incident radiation (angular dependence) were investigated. The average sensitivity reduction over the lifetime of the probes was 22.37% with a standard deviation of 0.63%. This reduction in sensitivity can be corrected for by the use of "drift equations". MOSFETs have a limited "lifetime" due to saturation effects with increasing accumulated dose. Saturation occurred at an average of 40 Gray (Gy) accumulated dose, for the high sensitivity probes investigated. The high sensitivity probes were linear within 1.6% for doses between 5 and 140 cGy, and 3.8% for the high sensitivity probes for doses between 50 and 500 cGy. Drift (changes in readings with time since irradiation due to electronic processes) over the long-term (from hours to weeks following irradiation) has been previously well characterised in the literature. This work focuses on shortterm drift, within the first few seconds or minutes following irradiation, being the most clinically relevant for in vivo measurements. Drift is investigated for various reading methods, such as reading frequency, and delays between irradiation and readings. It is shown that sensitivity, and consequently dose determination, is significantly influenced by the reading methodology. During the first five minutes following an irradiation, drift increased inversely with delivered dose, and was greater for probes having accumulated dose of > 20 Gy (2.0 -16.2% compared with 1.2 -7.4% for < 20 Gy probes). During the first five minutes following an irradiation, drift increased inversely with delivered dose, and was greater for probes having accumulated dose of > 20 Gy (2.0 -16.2% compared with 1.2 -7.4% for < 20 Gy probes). When two post-irradiation readings were taken following an irradiation, the difference between them generally increased as the time interval between the two readings increased, by up to 8.8%. Delays in taking pre-and post-irradiation readings resulted in drift of up to 5.7% or 9.3% respectively, compared with readings without a delay. These results emphasise the necessity for consistent methodologies between calibration and measurement in the clinical situation. Greater sensitivity was measured with the epoxy bubble, rather than the substrate side, facing the beam. The greatest variation, for orientations other than the bubble side facing directly towards the beam, was 10%, or 5% uncertainty in dose. The variations with angle were found to be reproducible, so that appropriate correction factors could be applied to correct measurements at angles other than with the sensitive area of the probes facing directly towards the radiation beam.