3D Printed Bolus With Flexible Materials: Treatment Planning Accuracy and Practical Aspects

Abstract

This work investigated the use of 3D printed flexible materials for use in constructing bolus in radiotherapy. Flexible 3D printed bolus can provide patients with a better treatment experience by improving conformity and comfort around delicate areas such as facial skin or ears. Previous studies have investigated hard 3D printed materials such as PLA or ABS, but this is the first look into flexible materials.Treatment planning system (TPS) accuracy was evaluated by comparing percentage depth dose (PDD) measurements made in a 3D printed phantom with those calculated by the TPS. Skin-bolus contact was assessed by CT scanning an anthropomorphic phantom with 3D printed bolus. Practical aspects of 3D printing with flexible materials were evaluated. The two flexible materials chosen for this study were Thermoplastic Polyurethanes (TPU): Ninjaflex and Cheetah (both from NinjaTek, Manheim, PA). The 3D bolus was printed on Fused Deposition Modeling (FDM) printers. 6x6 cm slabs of various thicknesses were designed in a 3D CAD program and 3D printed 100% solid. These were used to measure electron and photon PDDs from a depth of 0 to 2.0 cm, which are depths similar to those thicknesses typically employed with bolus. Photon and electron PDDs were measured with a plane parallel chamber and a diode detector, respectively. Both detectors were used in a holder printed in either Ninjaflex or Cheetah. Measured PDDs were compared to those calculated by a Pinnacle Enterprise TPS. This was done by applying radiotherapy beams to a CT scanned stack of 6x6 slabs and recording the dose at the various depths of interest. Practical aspects of constructing bolus with 3D printed flexible materials include printing speed, dimensional accuracy, specialized equipment and proper bolus-skin contact. Both materials printed to a dimensional accuracy within 0.2 mm. Ninjaflex being the more flexible material, requires a specialized extruder and also requires a slower printing speed (35 mm/sec) than Cheetah (80 mm/sec) in order to avoid extruder jams. Duration of print depends on print volume, but a small 3 mm thick nose bolus could take 3 hours with Ninjaflex and 1.5 hrs with Cheetah. Ninjaflex, however, demonstrates superior conformity due to its higher elasticity. The accuracy of the PDDs calculated by the TPS was within acceptable clinical tolerances. Heterogeneity correction was required to account for the 20% increase in density of the materials, compared to water. The Hounsfield units for Ninjaflex and Cheetah were measured to be -140 and 10, respectively, showing that Cheetah’s HU is similar to that of water. CT scans of skin-bolus contact showed areas where air gaps existed, which could be mitigated by applying US gel to the interface. Dosimetric properties and practical aspects of flexible materials for 3D printed bolus was investigated. These materials provide a high degree of conformity and comfort for patients. The Pinnacle TPS calculated PDDs accurately.