Degree Name

Doctor of Philosophy


School of Engineering Physics


Real time in vivo dosimetry is a major challenge in modern radiotherapy. One promising candidate for such dosimetry is the metal oxide semiconductor field effect transistor (MOSFET) dosimeter. The advantage of MOSFET dosimeters is that they offer real-time dose determination, along with a very small dosimetric volume. At the Centre for Medical Radiation Physics (CMRP) a new design of the MOSFET dosimeter called MOSkin, is being developed to measure the dose delivered to the basal layer (0.07 mm below the skin surface). This thesis discusses pre-clinical characterization of a new version of the MOSkin dosimeter and optimisation of the operating conditions.

Experiments were carried out to test a number of the characteristics of the dosimeter. The tests performed were to investigate the temperature response, sensitivity response, energy dependence and angular dependence of the dose response for the new MOSkin device. The effect of the packaging surrounding the MOSFET dosimeter in these properties was also investigated. Two versions of the MOSkin have been studied; one version was packaged “hard” (set in rigid packaging) and the other “soft” (no rigid packaging). The reproducibility of MOSkin measured dose was tested by placing a device at Dmax (15mm) in a block of solid water and acquiring data for multiple exposures with 10*10cm2 of 6MV photons. Each irradiation was of a set dose, with the dosimeter readout taken 10sec or 1 min after irradiation. Two phantoms were constructed, one plano-convex and one cylindrical, to test the angular response with the dosimeter located at the centre of each phantom. These were used to investigate the measured angular response of the dosimeter and the effect of phantom geometry on that response.

The temperature response tests of the un-irradiated MOSkin found that they have a thermostable readout current leading to low temperature instability 0.2mV/°C while sensitivity was 2.5 mV/Gy. In contrast to dual MOSFET dosimeters the single MOSkin includes built in thermo stabilization, which is independent of accumulated dose. Sensitivity response of the MOSkin was found to decline at a rate of 1.2% per 3Gy while creep-up effects were not observed in the MOSkin – waiting 10sec or waiting 1min after irradiation before readout made little difference on measured doses. The MOSkin sensitivity response like any Si device does depend on photon energy in the kV range with the response at 75kV producing 4.9 times greater than with 6MV photons. This is comparable to commercial MOSFET dosimeters described in literature with 4.5 higher than at 6MV for the same energy [33]. The angular dependence produced some interesting results. The back face of the hard device showed no angular dependence within ±2.5%, but there was angular dependence with the front face for ±90°. However the front face has been shown previously to be accurate for surface measurements when compared to the Attix ionisation chamber [12]. This presents the possibility for use of the MOSkin dosimeter both for surface and at depth in the current configuration.

The unique MOSkin design is showing to be very promising as a reliable real-time dosimetry device with the potential capability for many dosimetry applications, in particular skin dosimetry. Unlike other MOSFET dosimeters, it has been shown that the MOSkin has been designed to be temperature independent for dosimetry with a single device, while also having the unique capability of both surface and at depth dosimetry.