Investigation of MOSkin detectors for use in dosimetry of a HDR brachytherapy source

Dose verification is an important part of all radiotherapy. Dose measurements in high dose rate brachytherapy can be challenging for a number of reasons. The steep dose gradient around HDR sources necessitates a dosimeter with a very small detection volume. Also, the logistics of placing a dosimeter in a clinically meaningful position requires they have a small physical size. In this thesis we explore the use of two detectors for measurement of the backscatter factor for an application involving the treatment of colorectal cancer. These treatments involve excising the tumour then delivering a radiation dose directly to the tumour bed during surgery using an Intra-Operative Brachytherapy applicator. In these cases the radiation dose is sometimes delivered in the absence of backscatter material which can lead to a lower dose being delivered than predicted by the treatment planning computer.<br><br>Two dosimeters currently in use in brachytherapy are TLDs and MOSFETs. The use of TLDs in medical dosimetry is well established, they have a small physical size and an approximately tissue equivalent atomic number. MOSFETs have a less extensive history in medical dosimetry but their use is gaining popularity. MOSFETs have a small detection volume and are capable of giving readings in real time, making them ideal candidates for measurements around HDR brachytherapy sources.<br><br>A number of MOSkin detectors were provided for this project. The MOSkin is a variation on the MOSFET detector, the main difference being a lack of epoxy bubble covering the MOSFET chip. The uncertainty in a single measurement with a MOSkin detector was estimated by examining the following dosimetric characteristics – stability, accumulated dose response, linearity, angular response, energy response, dose rate response and temperature response. Similar characteristics were also examined for TLD rods in order to make a comparison. Measurements were made using a HDR Ir-192 source, a kilovoltage treatment unit and a high energy linac. Measurements were then made, using both TLD rods and MOSkins, to calculate the reduction in dose that results from a lack of backscatter material when using an IOBT applicator. <br><br>The largest variation in the response of the MOSkins was from accumulated dose, change in radial angle and radiation energy. For each of these characteristics the response was repeatable and therefore predicted by determining a correction factor from measurements. The largest variation in the response of the TLD rods was from changes in radiation energy which again was repeatable allowing the calculation of a correction factor. Using the IOBT applicator without adequate scatter material resulted in a reduction in delivered dose as measured with both the TLD rods and MOSkins and confirmed with previously published data calculated using Monte Carlo methods. MOSkin devices provide a useful measurement tool in the presence of high dose gradients, however, the dosimetric characteristics of the detector must be accounted for when estimating the uncertainty.<br>