IMPACT OF RSP ASSIGNMENT STRATEGIES ON DOSE CALCULATION ACCURACY FOR ABS 3D-PRINTED PATIENT-SPECIFIC COMPENSATORS IN PROTON RADIOTHERAPY

Background and Aims: Three-dimensionally printed, patientspecific compensators are increasingly being integrated into proton therapy workflows to improve dose conformity for shallow targets compared with treatment plans employing a standard nozzlemounted range shifter.In fused deposition modeling (FDM), the layer-by-layer process may introduce density variations along the printing axis, potentially affecting relative stopping power (RSP) representation in treatment planning systems (TPS) and proton range prediction.This study aimed to quantitatively evaluate the impact of different RSP assignment strategies on dose calculation accuracy using end-to-end experimental validation.Methods: Five clinical ABS compensators and corresponding reference cubes were analyzed to assess material homogeneity and temporal density stability.A dedicated ABS compensator was designed to investigate the dosimetric impact of four RSP assignment strategies: HU-and water-equivalent ratio (WER)-based, HU-based using the mean compensator HU, WER-based, and no RSP override using the clinical HU-RSP calibration curve.An end-to-end test including CT imaging, treatment planning, and dosimetric verification was performed.Dose distributions of monoenergetic proton fields were measured using a MatriXX detector.Gamma Index (GI) analysis (3%/2 mm) was performed in the plateau and distal fall-off regions.TPS-calculated dose planes were shifted along the beam direction within ±2 mm in 0.5 mm steps.Based on the GI results, depth-resolved maps were generated to determine, for each spatial location, the depth corresponding to the best agreement between calculated and measured dose distributions.Results: A reproducible HU gradient along the printing axis was observed, with HU variations of up to 3%, while the ABS material showed temporal HU stability.All RSP assignment strategies met clinical acceptance criteria (±3.5% proton range uncertainty).In the distal fall-off region, the no-override strategy achieved the best agreement with measurements.Uniform RSP assignment caused depth-dependent range inconsistencies, identified on depth-resolved maps as region-specific shifts of up to 1.5 mm (≈3.5% effective RSP deviation), compared to a maximum of 0.5 mm (≈1.25%) for the no-override approach.Conclusions: Uniform RSP assignment can obscure intrinsic FDM-induced heterogeneity in ABS 3D-printed compensators, leading to reduced stability in range-sensitive regions.Preserving CTderived HU values without RSP override retains spatial material information and provides the most robust agreement between TPS calculations and measurements.

Hali tarjima qilinmagan