ESTRO 2024 - Abstract Book

S3687

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2024

and uses free script based CAD-software OpenSCAD as backend for the 3D file construction and the free Python package Tkinter to construct the graphical user interface. The user interface allowed the selection of the DICOM plan file from the Therapy Planning System (TPS). We have tested the file format for Pinnacle 16.2 and Varian Eclipse 15.6 and 18. Block and aperture geometry requirements conform to ELEKTA Synergy and VersaHD linacs. A point reduction algorithm was implemented to reduce unnecessary print head movement. Preview windows were provided for the user to check the plausibility of the geometry and point reduction. Cura 5.4 prepared the complete toolpath for the Fused Deposition Modeling (FDM) printer to produce objects. Parameter sets and heat resistant materials were evaluated to print reliable moulds for electron apertures and photon blocks in external beam radiotherapy - in particular for TBI total body irradiation. The process and parameters were determined for the ULTIMAKER S5 printer - a representative of the widely used and inexpensive FDM printers. The dimensional accuracy of the cast blocks and apertures was evaluated in comparison to the planned dimensions and in comparison to those produced with the hot wire cutter. We demonstrate the applicability of low-cost FDM printers. We have developed an open source software tool to convert block or aperture beam data from DICOM-RT plan files into templates for 3D printable moulds or spacers. The accuracy of the inexpensive FDM printers is superior to computer-controlled hot-wire machines, which burn a gap of estimated1-2 mm in the polystyrene with a given accuracy specification of +-1mm (BEST, Compu-Cutter III). Hand guided cutters, can easily burn even larger gaps, especially on small curves or edges due to speed changes and wire deflection; the accuracy of a straight line depends on the skill and experience of the user. The accuracy in the printed shapes is +- 0.3mm for the square (50x50mm), cylindrical (50mm) and plus-shaped (100x100mm) test aperture and test block. After establishing reliable printing parameters and material properties, we could not detect shrinkage or visible bulging effects in the cast metal parts. Minimal wall width for the used material was 3.2mm (4 lines with 0.8mm nozzle) to not bulge under heat and stress of the casting process for photon blocks. As sufficiently heat resistant printing material we found Extrudr Greentech Pro with a heat resistance of 160°C. Ultimaker Cura can calculate moulds which was used for photon blocks. Electron aperture spacers are objects with 2.4mm wall width, no top or bottom layers and a light 7% infill grid. All parameters are stored in parameter sets in Cura. The hands-on time for the user is comparable to the old method with the computer-assisted hot wire cutting machine. The only drawback we could find was the time taken to print. Print time for a typical TBI on floor - lung electron aperture (13mm high) is about 1h per spacer. Print time for a typical 80mm high photon lung block mould is about 2h. Typically we group 4 spacers on one print bed and 4 blocks on another print bed. Results:

Conclusion:

Although the hands-on time of 3D printing is comparable to that of hot wire cutting, the time required to print is a disadvantage. If workflow and schedule permit, overnight printing is recommended. As a backup and to increase capacity and output, our department uses two identical printers. The 3D printed individual and divergent electron apertures and photon blocks proved to be suitable for everyday clinical use, even with high precision requirements.

Keywords: FDM, software, collimation