In-Vivo Dosimetry for HDR Prostate Brachytherapy

Abstract

The main challenge in radiotherapy is to accurately deliver a prescribed radiation dose to the tumour while limiting the dose to surrounding healthy organs and tissues. The ultimate quality assurance would be the use of an online in-vivo dosimetry system, which would verify the dose delivered to the patient in real-time while treatment is delivered. Furthermore, in-vivo dosimetry measurements could be compared to the treatment planning system. The dosimeter used for this study was the Metal-oxide semiconductor field-effect transistor because of its several advantages. The main objective was to develop a real-time in-vivo dosimeter service for High Dose Rate prostate brachytherapy. This study involved commissioning and calibrating the dosimeter for use in High Dose Rate prostate brachytherapy. Tests included linearity, calibration, consistency and reproducibility, anisotropy (azimuthal and polar), temperature dependence and energy dependence. Six treatment plans created on a phantom meeting clinical guidelines were created using the treatment planning system under ultrasound guidance and these were used to first generate the correction factors required to produce the predicted voltage to the dosimeter, and finally assessing three of the six plans in real-time. The dosimeter responds linearly with dose, over the clinical dose range (0.01 Gy to 20 Gy), with R2= 0.9991. Anisotropy azimuthal and polar resulted in a minimal dependence within the uncertainty of the measurements; hence no azimuthal and polar correction factors were applied to phantom measurements. The dosimeter is energy dependent hence; a distance dependent energy correction factor was applied to treatment planning system values to produce the predicted value. An average of the calibration correction factors was also applied to the treatment planning system values to convert the dose (Gy) to (mV). Temperature was found to be dependent with a percentage difference of 7% between calibration temperature and body temperature. The measured phantom plans (mV) were in good agreement with the predicted voltage of the dosimeter (mV). The maximum percentage differences between the measured and the predicted was of 5.082%. Clinical plans in real-time were analysed and found that this is an optimal way to detect gross errors per individual needles. Total uncertainty budget of this study was 9.97% for k=2. In-vivo dosimetry in brachytherapy is currently not well established in clinical routine use to detect errors; hence, implementing this dosimeter is the way forward to start verifying treatments and therefore being capable of detecting gross errors.