BackgroundComputed tomography (CT) data used for patient radiotherapy planning can nowadays be used to create 3D-printed boluses. Nevertheless, this methodology requires a second CT scan and planning process when immobilization masks are used in order to fit the bolus under it for treatment.This study investigates the use of a high-grade surface-scanner to produce, prior to the planning CT scan, a 3D-printed bolus in order to increase the workflow efficiency, improve treatment quality and avoid extra radiation dose to the patient.MethodsThe scanner capabilities were tested on a phantom and on volunteers. A phantom was used to produce boluses in the orbital region either from CT data (resolution ≈1 mm), or from surface-scanner images (resolution 0.05 mm). Several 3D-printing techniques and materials were tested. To quantify which boluses fit best, they were placed on the phantom and scanned by CT. Hounsfield Unit (HU) profiles were traced perpendicular to the phantom’s surface. The minimum HU in the profiles was compared to the HU values for calibrated air-gaps. Boluses were then created from surface images of volunteers to verify the feasibility of surface-scanner use in-vivo.ResultsPhantom based tests showed a better fit of boluses modeled from surface-scanner than from CT data. Maximum bolus-to-skin air gaps were 1-2 mm using CT models and always < 0.6 mm using surface-scanner models. Tests on volunteers showed good and comfortable fit of boluses produced from surface-scanner images acquired in 0.6 to 7 min. Even in complex surface regions of the body such as ears and fingers, the high-resolution surface-scanner was able to acquire good models. A breast bolus model generated from images acquired in deep inspiration breath hold was also successful. None of the 3D-printed bolus using surface-scanner models required enlarging or shrinking of the initial model acquired in-vivo.ConclusionsRegardless of the material or printing technique, 3D-printed boluses created from high-resolution surface-scanner images proved to be superior in fitting compared to boluses created from CT data. Tests on volunteers were promising, indicating the possibility to improve overall radiotherapy treatments, primarily for megavoltage X-rays, using bolus modeled from a high-resolution surface-scanner even in regions of complex surface anatomy.

No abstract

30 cm by 30 cm scoring volume with 1 mm voxels was used to estimate the dose in the cavity and surrounding tissue. Results: The depth dose calculation for single dwell position with Gean4 simulation in homogenous water phantom shows good agreement with Oncentra TPS dose calculation. The simulated SAVI 6-1 applicator, on the other hand, showed a 12 % higher dose at the tissue lung interface compared to the TG-43 dose calculation. Also the simulation showed an overestimated of 5% dose on the skin depending on the location of the air cavity. Conclusion: The dosimetric effect of air cavity highly depends on the cavity size, dwell position and dwell times. The Geant4 Monte Carlo simulation estimated the dose increase on tissue-lung interface and on skin in the range of 5-12 % depending on the location of the cavity, the dwell positions of the sources and the dwell times of the those sources.