Stereotactic ablative radiation therapy (SABR) involves the precise delivery of high doses of radiation to target volumes. Such treatments require high levels of confidence in their delivery due to the high doses and conformal beams, where an inaccurate delivery could have major adverse effects. To provide confidence in the delivery of such treatments, phantoms are used as a quality assurance (QA) tool to verify the delivered dose is as predicted by the treatment planning system (TPS). Commercial phantoms are available, however they are limited in their function and come at a large cost to the clinic, meaning it is unlikely for such a phantom to be sent between centres given the negative implications in the event of damage.

In this thesis, an anthropomorphic phantom fit for dosimetry verification measurements across the New Zealand radiation oncology departments was designed, fabricated, and verified in its clinical use in generating inter-department dosimetry comparisons to provide confidence in lung and spine SABR treatments. The final phantom was encased in an acrylic sheet and 3D printed PLA. It consisted of polyurethane, an epoxy resin and phenolic microsphere (PMS) compound, epoxy resin, and Teflon as equivalent materials for soft tissue, lung, heart, and bone respectively. CT analysis, Monte Carlo simulations, and investigations into their physical and chemical properties confirmed that materials were relatively radiologically equivalent to such tissues. Tumours were created with various sizes and densities using a polyurethane and PMS compound to allow for adaptive radiation therapy verification capabilities. Tumours within the lung equivalent material undergo simulated respiratory motion through simple sinusoidal motion, or will iterate through motion tracer files acquired from patients undergoing 4DCT.

Verification of the constructed phantom was performed at the St George’s Cancer Care Centre (SGCCC) where multiple CT scans were taken using clinical thorax and physics protocols to gain a clinical perspective of the phantom and to provide higher resolution images respectively. Through CT images, the phantom was confirmed to be created of relatively homogeneous materials, with air bubbles only occurring at the surface of each body portion. Materials were confirmed to have approximate tissue equivalence through their CT numbers and relative electron densities compared to ICRU recommendations and patient data. A complex SABR plan created for patient use was delivered to the lung, heart, and spine, and a simple plan with a larger field size was delivered to the spine for a more uniform dose distribution to better visualise dose fall off to nearby structures.

Transmission measurements were performed to verify the agreement of dose between the pre- scribed and delivered plans. For dose to medium (DTM) TPS calculations, ionisation chamber measurements recorded differences of 1.14%, -0.55%, and 1.66% for doses delivered to the heart, spine, and spinal cord when applying relevant shifts to account for positioning errors during delivery, all within the 2% recommended tolerance, as per SGCCC clinical process. Alongside this, all film measurements taken within the lung equivalent material produced appropriate results within the 95% confidence rate for gamma analysis when applying a 2%/2 mm gamma criterion using red channel analysis. Pass rates of 98.60%, 99.24%, and 100% were achieved for stationary lung tumours with densities of approximately 0.567, 0.872, and 1.023 g/cm3 respectively. The pass rates for the high and low density tumours decrease when planned to a medium density tumour, indicating the need for adaptive capability in the phantom. Film measurements taken within the spine showed potential positioning errors when applying the simple plan. Utilising the same gamma criterion and film analysis, a pass rate of 56.28% was achieved, which was improved through scaling to 84.34%. In green channel analysis, as used in the analysis of high doses, a pass rate of 90.18% was achieved, however this is still out of tolerance for such a measurement. Therefore, the phantom was deemed suitable for dosimetry verification measurements within the lung and heart, however further analysis is required prior to clinical use for its spine treatment verification.