An Anthropomorphic Multimodality (CT/MRI) Head and Neck Phantom for Radiotherapy Applications

Aims:
To develop and evaluate an anthropomorphic multimodality phantom for the head and neck (H&N) anatomy that can be used with computed tomography (CT) and magnetic resonance imaging (MRI) for radiotherapy (RT) applications. The research aims to identify suitable materials for creating these phantoms, assess the suitability and effectiveness of a 3D head and neck phantom for MRI-based quality assurance (QA) in RT planning, and to optimise cone beam computed tomography (CBCT) protocols for H&N imaging as part of QA processes.

Methods:
Through literature research, candidate materials potentially suitable for developing multimodality (MRI/CT) phantoms were identified and produced. Their suitability and stability over time and after exposure to radiation were then evaluated. An anthropomorphic multimodality H&N phantom was used to evaluate the benefits of using such a phantom for conducting QA tests recommended by international bodies in MRI guided RT treatment planning services. Moreover, the scope of the phantom's use has been expanded to include optimising CBCT protocols, further demonstrating its value in enhancing QA processes across multiple imaging modalities.

Results:
The results of this project indicate that while some materials meet specific requirements for creating anthropomorphic multimodality phantoms, it has been challenging to find materials that simultaneously satisfy the needs of both MRI and CT modalities. However, the results have shown that the T1 and T2 relaxation times and CT numbers of 10% polyvinyl alcohol cryogel closely match those of normal brain grey matter, and remain stable over a year, and after exposure to radiation levels up to 1000 Gy, demonstrating its potential effectiveness in making phantoms. The anthropomorphic multimodality phantom has demonstrated superior performance to non-anthropomorphic phantoms in certain aspects of MRI-based RT planning QA, particularly in end-to-end testing. The phantom can be used in optimising CBCT protocols as part of QA processes, with results showing that it allows for a reduction in radiation doses by more than 50% compared to the default protocol for patients with head and neck tumours without significantly affecting image or registration quality and with the expectation that this would not have a consequential impact on treatment plans.

Conclusions:
The identification of only one suitable material underscores the need for expanded research into multimodality phantom materials. The phantom proves effective for MRI-based QA. Additionally, it was employed to test and optimise CBCT protocols, leading to reductions in radiation doses without compromising image quality.