Camera-based radiotherapy dosimetry using dual-material 3D printed scintillator arrays

To describe a methodology for the dual-material fused deposition modeling (FDM) 3D printing of plastic scintillator arrays, to characterize their light output under irradiation using an sCMOS camera, and to establish a methodology for the dosimetric calibration of planar array geometries.

We have published an investigation into the fabrication and characterization of single element FDM printed scintillators intending to produce customizable dosimeters for radiation therapy applications. ¹ This work builds on previous investigations by extending the concept to the production of a high-resolution (scintillating element size 3 × 3 × 3 mm³ ) planar scintillator array. The array was fabricated using a BCN3D Epsilon W27 3D printer and composed of polylactic acid (PLA) filament and BCF-10 plastic scintillator. The array's response was initially characterized using a 20 × 20 cm² 6 MV photon field with a source-to-surface (SSD) distance of 100 cm and the beam incident on the top of the array. The light signals emitted under irradiation were imaged using 200 ms exposures from a sCMOS camera positioned at the foot of the treatment couch (210 cm from the array). The collected images were then processed using a purpose-built software to correct known optical artefacts and determine the light output for each scintillating element. The light output was then corrected for element sensitivity and calibrated to dose using Monte Carlo simulations of the array and irradiation geometry based on the array's digital 3D print model. To assess the accuracy of the array calibration both a 3D beam and a clinical VMAT plan were delivered. Dose measurements using the calibrated array were then compared to EBT3 GAFChromic film and OSLD measurements, as well as Monte Carlo simulations and TPS calculations.

Our results establish the feasibility of dual-material 3D printing for the fabrication of custom plastic scintillator arrays. Assessment of the 3D printed scintillators response across each row of the array demonstrated a nonuniform response with an average percentage deviation from the mean of 2.1% ± 2.8%. This remains consistent with our previous work on individual 3D printed scintillators which showed an average difference of 2.3% and a maximum of 4.0% between identically printed scintillators.¹ Array dose measurements performed following calibration indicate difficulty in differentiating the scintillator response from ambient background light contamination at low doses (<20-25 cGy) and dose rates (≤100 MU/min). However, when analysis was restricted to exclude dose values less than 10% of the Monte Carlo simulated max dose the average absolute percentage dose difference between Monte Carlo simulation and array measurement was 5.3% ± 4.8% for the fixed beam delivery and 5.4% ± 5.2% for the VMAT delivery CONCLUSION: In this study, we developed and characterized a 3D printed array of plastic scintillators and demonstrated a methodology for the dosimetric calibration of a simple array geometry.