Degree Name

Master of Research


School of Physics


This thesis aims to characterise the accuracy of a new active CMOS pad detector for dosimetry in HDR Brachytherapy. A 1x1x1 mm3 detector is doped with high resistivity n-type silicon, thickness of 100 μm (Wafer10) and 48 μm (Wafer20) epitaxial substrate and built in standard CMOS along with a transimpedance amplifier. This work investigates and examine this detector’s sensitivity and signal to noise ratio in air for use in High Dose Rate Brachytherapy (HDR BT) dosimetry.

To understand operational characteristics of the detector, a current-voltage (I-V) characteristic curve was created, and the depletion voltage of the junction observed by using a Keithley Programmable Voltage Source to sweep the voltage and measure the current collected at the pad. Applying the established depleted voltage, the detector sensitive volume was tested by irradiating 5.5 MeV alpha beam with field size 1 mm x 1 mm at 1000 Hz event rate to observe the impact of radiation on the electronic and collection efficiency of the detectors. Source tracking was performed in the clinic by dwelling HDR Ir-192 Flexisource (Elekta, Veenendaal, the Netherlands) at different depths and using the DDC226 interface to track the charge absorption. The sensitivity and responsiveness were assessed and a conversion factor in nC/Gy was calculated for use in further data analysis.

The minimum voltage found to produce full charge collection efficiency is −30 ± 0.1𝑉 for both Wafer10 and Wafer20. A HDR brachytherapy Ir-192 Flexisource (Elekta, Veenendaal, the Netherlands) was used for source linearly measurements and a dose comparison were made to AAPM TG-43. A variation of within 5% between measured absorbed dose and TG-43 was found corresponding to the difference of +/-2cGy. Higher dose differences were measured for 2 mm and 7 mm depths due to scattering and radiation field perturbation created by the ceramic/golden package adopted for assembling the sensor prototypes. The accuracy in reproducing dwell positions lies within 1 mm compared to actual distance set by TPS. The sensitivity found for Wafer10 is (5.8 ± 0.021) x 10-8 and (3.6 ± 0.14) x 10-8 for Wafer20. This supports the correlation between sensitivity and the material’s electronical and architectural design.

The response of the detector when irradiated using HDR Ir-192 placed in water equivalent phantom obtained within 0.1% of the time set by TPS. Due to the structure, design, and air gap within the space of the sensor region, scattering and field perturbation prevents the gamma reaching the sensor resulting in high undervaluation of variation in the depth dose response compared to TG-43 for 2mm and 7mm. But for larger depths of 12 to 52mm, the variation is within 5% compared to TPS corresponds to the difference of +/-2 cGy. Further investigation of angular dependence and field perturbation would be beneficial to confirm the proposed detector is suitable for clinical applications such as quality assurance and dosimetry in HDRB.

FoR codes (2008)

029903 Medical Physics



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.