The Effect of Filament Density on the Build-Up Region Responsiveness of 3D Printed Boluses in High-Energy Photon Radiotherapy

Abstract

Due to the inability of commercial boluses to fully replicate the body's curvature during radiation treatments, introduction of 3D printed boluses can be a solution for avoiding the skin sparing effect for irregular body shapes. The purpose of this study was to assess the applicability of a uniformly thick 3D printed bolus. We report the influence of 6 different 3D printed boluses on the surface dose (Ds) delivered by a 6 MV photon beam linear accelerator. Six distinct types of filaments used in 3D modelling were evaluated. The selection of materials was determined by the filament's density, to be comparable with a commercial bolus used in clinical practice (tissue equivalent). Bolus plates of 21.5x21.5x0.5cm were designed, printed and exposed to high energy photon beam with field apertures of 5x5cm, 10x10cm, 15x15cm, and 20x20cm. We used a water tank phantom, which gives precise beam data visuals to confirm the delivered dose. The behavior of the profile depth dose (PDD) was analyzed for different air volumes between the bolus surface and water characterized by bolus-water distances of 0cm, 0.5cm, 1cm, and 1.5cm. A conventional beam geometry with SSD = 100cm and scanning depths of 15, 50, 100, 200, and 300 mm was used. The initial validation of the materials was performed by evaluating the CT attenuation properties (mean Hounsfield Units - HU) of the six 3D printed boluses and the commercial one. The commercial bolus had -34.89HU, whereas the ABS (-144.53HU), ASA (-124.40HU), Premium PLA (9.55HU), Polycarbonate (-140.79HU), Standard PLA (-68.58HU) and PET-G (-113.159HU) had comparable attenuation values. The PDD reference measurements without bolus resulted in the following surface dose (Ds) values: 54.42% for the 5x5cm field, 57.94% for 10x10cm, 61.29% for 15x15cm, and 65.44% for 20x20cm, respectively. We observed a proportional increase of Ds with the field size. Even more, when the bolus was in direct contact with water (0cm airgap), an influence of the material density on the Ds values was evidenced. For the commercial bolus (ρ = 1.01) we obtained a Ds = 95.59%, while ABS (ρ = 1.04) had Ds = 89.12%, ASA (ρ = 1.07) Ds = 86.51%, Premium PLA (ρ = 1.17) Ds = 90.89%, Polycarbonate (ρ = 1.19) Ds = 88.55%, Standard PLA (ρ = 1.24) Ds = 89.15%, and PET-G (ρ = 1.27) Ds = 88.78%, respectively. The Ds value dropped following the increase of the introduced air volume. The Ds values and the analysis of the buildup regions in the depth dose profile revealed the significance of the air present between the skin surface and the bolus in the clinical setting, as well as the need to reduce these air gaps. We confirmed from a dosimetric perspective, that the materials used for the 3D printed bolus plates are suitable to be used as a clinical bolus. Further validation studies are warranted to exploit the potential of these 3D printed boluses in real clinical scenarios.