Degree Name

Master of Philosophy


School of Physics


Background: Image-guided radiation therapy (IGRT) has drastically enhanced the accuracy of radiation delivery and has developed as the new worldview for patient positioning and target localization in radiotherapy. Using a Kilo Voltage Cone Beam Computed Tomography (KV CBCT) approach, three dimensional (3D) volumetric patient anatomy is available for patient positioning for each fraction. This helps in adaptive planning as seen in newly developed scenarios that contributes to better care for patients.

Although the addition of the kV CBCT imaging modality into radiotherapy increases treatment precision, it also increases the patient dose without it being displayed on the planned dose. Separately estimating the CBCT dose is considered important to achieve an accurate planned dose to avoid patient doses exceeding critical organ threshold levels. In most of the planning systems however, the effect of CBCT dose is neglected as it is thought to be negligible in comparison with Mega Voltage (MV) therapy dose.

Aim: This project aims to determine the patient doses from CBCT during a selected number of radiotherapy procedures. This is achieved through direct measurement radiation dose in a phantom and through mathematical modelling utilising existing software used in diagnostic radiology dosimetry. A second aim then is to compare these methodologies.

Methods: The evaluation of CBCT doses was achieved through two different methodologies. In the first method Thermoluminescent Dosimetry (TLD) measurements were performed using a Rando phantom on a VarianTM On Board Imager (OBI) Linac system utilising manufacturer specified CBCT protocol parameters for Standard Head and Pelvis Spot CBCT protocols. For this project TLD measured values were regarded as the base line values. In the second method, the Monte Carlo code PCXMCTM was used in rotation mode utilising ExcelTM. Additional modelling was required due to the fact that PCXMC uses a flat filter while CBCT units utilise shaped filters such as the Bow-tie filter. This involves the determination of procedure protocols and dosimetric parameters associated with each procedure investigated. For the units examined at Canberra Hospital the manufacturer specified the CBCT protocol parameters. Dosimetric parameters included quantifying the effect of the bow tie filter and general beam output. The effect of patient size was investigated as dose-modifiers through changes in the PCXMC patient model, including the relation between specific critical organ doses and patient size. Finally, the organ doses obtained through each of the two methodologies were compared.

Results: The CBCT doses for critical organs identified by the standard head and pelvis spot protocols were measured using TLDs and calculated through PCXMCTM simulations. For a full radical brain series of treatment fractions, the highest estimated total organ dose is 180 mGy for the parotid gland. For a full bladder series of treatment fractions, the highest estimated total organ dose is 230 mGy in the uterus. For a full intact prostate series of treatment fractions, the highest estimated total organ dose is 810 mGy.

Comparing the TLD measured results and PCXMCTM simulated results, it can be seen that the dose to the parotid gland matches within 10%, whereas the spinal cord dose comparison shows almost a 30% difference. It is also to be noted that certain organs cannot be properly compared due to the absence of holes in some of the actual organ positions in the phantom.

The comparison of PCXMCTM software simulated results for standard Rando man phantom and standard Australian man phantom shows a consistent variation between organ doses ranging from 5% to 18 % with lower doses seen in the standard Australian man. This highlights the fact that as the patient size reduces dose received by the patient increases for a given set of exposure parameters.

Discussion: When estimating the total doses due to CBCT protocols, difficulties were encountered in that the number of radiotherapy fractions for a given radiotherapy treatment regime varied greatly. The total dose contribution due to the CBCT protocol is currently ignored in the calculation of the clinical radiotherapy treatment dose. The methodology developed and utilised in this thesis shows that there is a small, but measurable, contribution to the organ dose from the CBCT protocols performed during a typical radiotherapy treatment regime. The highest organ dose was for the uterus in pelvic-related radiotherapy treatment where up to 6% of the threshold dose might be achieved through the use of CBCT alone.

The patient size calculations demonstrated that with the current use of fixed radiographic factors, as the size of the patient increases the total dose to critical organs reduces. However, the increasing size of the patient will simultaneously decrease the quality of the image when measured in terms of the contrast to noise ratio. Similarly, if such images are clinically adequate then implementing variable radiographic factors would see a corresponding decrease in the total dose received by smaller patients.

Conclusion: The results presented in this thesis demonstrate that there is a small, yet significant, total dose contribution due to the current CBCT protocols typically used as part of radiotherapy treatments in cancer clinic today. It was shown that in some cases the additional dose should be considered in the treatment planning process due to possible impacts on organ threshold values. The additional CBCT dose will of course, add to the increased risk of secondary cancers, but this is a very minor impact.

The results also indicate that the radiation dose varies significantly with both patient size and tumour position for both standard head and pelvic spot scanning protocols. Therefore, a recommendation arising out of this thesis work is that the current practice of using fixed scan parameters, independent of patient size, be reconsidered to incorporate patient-specific imaging protocols. Especially for the case of paediatric patients who will otherwise receive a higher CBCT related dose.



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.