ESTRO 2024 - Abstract Book

S3174

Physics - Detectors, dose measurement and phantoms

ESTRO 2024

Despite the known limitation of the calculation accuracy of spatial dose distributions due to density differences, phantoms with homogeneous density are used in clinical practice. This is mainly due to the lack of availability of inhomogeneous phantoms for plan-specific QA. Here, we describe the creation of a vertebral phantom with anatomy like inhomogeneities using a 3D printing process. The phantom was tested for its usefulness in plan-specific QA.

Material/Methods:

Print parameters were determined that allow to reproduce bone density. Test specimens with varying infill parameters were produced using StoneFil (Formfutura® BV) filament in an UltiMaker S5 printer using a fused deposition modeling (FDM) process. These specimens were created with either a sealed wall to retain air or permeable wall allowing water to enter the infill structure. CT scans of the specimens submerged in water were acquired using a SOMATOM go.Open Pro. Density was derived from the images. Subsequently, 3D-printed vertebrae replicas were fabricated. Lumbar vertebrae L1, L2 and L3 were segmented from a publicly available CT dataset in the Pinnacle treatment planning system (TPS). 3D models were exported, and their surface was smoothed using Meshmixer before 3D-printing. A planning CT was acquired with the vertebrae positioned inside a water-filled plastic barrel of 32 cm in diameter. Twenty treatment plans of varying complexity were created for this phantom. These plans were computed in the TPS using the Collapsed Cone Convolution algorithm. Subsequently, these plans were compared to measurements on a Versa HD (Elekta) at 6 MV using an ion chamber (PinPoint 3D 31022), unshielded diode (E-Type 60012) and a microDiamond (60019, PTW Dosimetry). Reference measurements were performed using a scintillator (W2-1×1, Standard Imaging). They were all positioned in the foramen vertebrale, serving as an estimate for the dose to the spinal canal. CBCT images in the irradiation position were obtained and added to the TPS as secondary datasets to obtain exact position of the detectors’ measurement volumes. Additional plan-specific quality assurance using a homogeneous ArcCHECK phantom was performed.

Results:

StoneFil with 20 % infill density and an undulating pattern that changes directions and is completely permeable to water and finally filled, create bone marrow equivalent density in the CT (Figure 1). Residual air gaps were eliminated by immersing the printed objects in water. The cortical structure of the vertebral bodies is determined by the print parameters of the wall thickness. After considering CT images, a wall thickness of 2 mm was deemed suitable.