Additive manufacturing (AM) enables the low-cost and patient-specific manufacture of anthropomorphic Radiation Dosimetry Phantoms (RDPs), used for the pre-treatment planning of cancer patients, to validate target doses and minimise the ionising radiation effects towards adjacent healthy tissues. By Rance Tino, Darpan Shidid, Bill Lozanovski, David Downing, Martin Leary, Tomas Kron and Milan Brandt of the RMIT Centre for Additive Manufacturing.
Radiotherapy aims to deliver a curable radiation dose to tumours while sparing surrounding healthy tissue, which is achieved by the accurate conformal delivery of ionising radiation via an external beam using linear accelerators, or an internal beam using sealed radiation source (called brachytherapy). Modern radiotherapy involves CT-simulation, 3D-treatment planning and its quality assurance processes prior to patient treatment to produce highly conformal dose distributions and to ensure its safe and accurate delivery.
It is common to build anthropomorphic RDPs through moulding and casting, to mimic the radiation properties of humans as a radiation dose cannot be directly measured in patients. As part of quality assurance (QA) of patient treatment plans, patient-specific dose measurements are often performed using RDPs combined with various dose measurement tools.
Unfortunately, anthropomorphic RDPs manufactured through traditional moulding and casting techniques are associated with high fabrication costs and long processing times. In addition to this, they are not patient-specific in terms of individual dimensions (particularly in respect to obese patients), feature standardised tissue heterogeneity, and lack pathological features.
This article discusses some of the basic concepts surrounding the manufacture of AM-RDPs, their clinical significance and requirements and their associated printing techniques and materials utilised at the RMIT Centre for Additive Manufacturing.
Anthropomorphic RDPs for treatment planning
The treatment planning procedure is a significant part of radiotherapy, whereby the optimal treatment parameters to be used for the management of a patient’s disease are determined. These treatment parameters include target volume, dose-limiting structures, treatment volume, dose prescription, dose fractionation, dose distribution, the positioning of the patient, treatment machine settings, adjuvant therapies.
The role of commercially available anthropomorphic RDPs is to act as a human proxy making it possible to experimentally visualise and evaluate treatment options tailored to the locality of patients. This importance signifies the current limitations of anthropomorphic phantoms as they only follow the average radiation and body dimensions of a ‘healthy’ person, with a lack of patient-specific pathological features – in particular, the mimicry of accurate lesion size and positioning. Therefore, research opportunities exist for AM technology in highlighting these limitations due to its conformal and rapid prototyping capabilities.
Patient-specific radiotherapy phantoms enabled by AM
AM provides opportunities for the inexpensive manufacture of patient-specific devices, as observed from the current literature not only for radiotherapy phantoms but also for other radiotherapy devices such as bolus, compensators, electron beam shielding, immobilisers, and brachytherapy moulds.
Novel AM workflows have been developed to accommodate imaging tissue-like heterogeneity utilising AM materials, via modification of infill parameters, doping, and the introduction of voided geometric features as the structural basis for AM-RDPs.
Types of AM radiotherapy phantoms
Early versions of additively manufactured radiotherapy phantoms were manufactured as shell phantoms, which are hollowed phantoms filled with various tissue-equivalent materials (such as sawdust, silicone gels, or cork). The emergence of better AM technologies has attracted interest in exploring the simulation of the human tissue heterogeneity, classified as as-printed phantoms.
Heterogeneity in printed phantoms can be achieved using modified material extrusion (fused deposition modelling (FDM)) printing parameters such as infilling patterns and percentage, printing nozzle size, temperature, and more recently, the modification of material extrusion rate using the Pixel-by-Pixel (PbP) method. Furthermore, contrast variations can also be achieved by constructing phantoms with two or more different AM materials (multiple material printing); doping filaments with high-density materials such as bismuth and barium sulphate to increase the observed HU range; and the use of controlled voided structures within the manufactured phantoms to precisely controlled HU values.
Recent studies have illustrated the combination of these manufactured phantoms with commercially available motion platforms and in-house motion devices to further simulate body movements, especially the thorax’s respiratory movements (classified as 4D-AM phantoms).
Clinical requirements and implications
Recently, printing guidelines and recommendations for manufacturing AM-radiotherapy devices have been developed by the SIG (Special Interest Group on 3D printing), a writing group representing the Radiological Society of North America. They are divided into four main processes including:
- Medical image acquisition – Commonly used imaging modality involves CT or MRI. Associated patient data should have sufficient spatial resolution to accurately represent anatomy to be modelled.
- Image data preparation and manipulation – This includes image segmentation, 3D CAD design, and file documentation.
- Generation of the 3D-printed model – This involves the printing process, post-processing, and model inspection.
- Quality Control program – This involves the delivery and discussion with referring physicians, pre-operative planning, material biocompatibility, cleaning and sterilisation, and clinical appropriateness.
Regarding printing materials, it is essential to consider the photoelectric and Compton effects when comparing result outputs with human tissues. Photoelectric effect serves as the dominant phenomena at low X-ray energies ranging below 200KeV, hence for imaging modalities (CT, MRI, PET). At higher X-ray energies up to 10MeV, Compton effects can be considered as the dominant phenomena, where material attenuation differs depending on their elemental composition, signifying how radiation doses are distributed. Ideally, additive manufactured RDPs aim to simulate not only the patient’s proportion and pathological features but also the imaging attenuation of human tissues, the photoelectric effect, the dose attenuation of tissues, and the Compton effect.
Also, for given printing material to be tissue or water-equivalent, it must have the same effective atomic number, number of electrons per gram, and mass density. However, since the Compton effect is the most predominant mode of interaction for MV photon beams in the clinical range, the necessary condition for water equivalence for such beams is the same electron density (number of electrons per cubic centimetre) as that of water.
AM playing with radiotherapy? Or radiotherapy playing with AM?
Despite the enabled low-cost and patient-specificity of AM radiotherapy phantoms, the associated printing techniques and materials are limited and are yet to be converged in terms of reproducibility, where manufacturability issues of current printing technologies are still present. FDM technology, in particular, is commonly used for manufacturing radiotherapy phantoms. This printing technique comes with inherent limitations in comparison with other printing techniques such as polymer jet printing and stereolithography (SLA), where observed void defects are observed, which in turn produces structurally weak and non-uniform dense objects.
Researchers at RMIT University are currently investigating the manufacturing process of these radiotherapy phantoms and exploring how they can be used in a clinical setting considering the required manufacturing compatibility, accuracy, time and cost. In highlighting previously mentioned manufacturing limitations, a unique geometrical structure called a ‘Gyroid’ is also being investigated by these researchers as they enable: controllable printing tool path parameters in minimising void defects; controllable porosity at all directions highlighting tissue-like heterogeneity and offering a similar tissue-like structure for assessing tissue deformability.
The RMIT Centre for Additive Manufacturing is involved in research projects with collaborators at the Peter MacCallum Cancer Centre, Melbourne, at Stryker South Pacific, St Vincent’s Hospital, UTS, IMCRC, DMTC, DSTG, QUT, University of Wollongong, Swinburne University, Ford and RUAG. This research is supported by the ARC Training Centre in Additive Biomanufacturing, which focuses on the research & development of new biomedical products using AM technology.