ESTRO 2021 Abstract Book

S1360

ESTRO 2021

Conclusion Optimization of the initial model allowed an improvement of at least 2.5% for representative plans of the clinical practice of the center, demonstrating that the modeling of the MLC is the most critical part to achieve the required precision. PO-1639 Commissioning and clinical use of a measurement-guided dose reconstruction (MGDR) for Tomotherapy P. Sánchez-Rubio 1 , A. Montes-Uruén 1 , R. Rodríguez-Romero 1 , N. Gómez-González 1 , J. Martínez-Ortega 1 , M. Torres-López 1 1 Hospital Universitario Puerta de Hierro Majadahonda, Medical Physics, Majadahonda, Spain Purpose or Objective Since the gamma method has proven to be ineffective in detecting dosimetric errors, commercial softwares have been developed to evaluate the feasibility of the plan in terms of DVH metrics. In this work, a validation method for the ArcCHECK® + 3DVH® for Tomotherapy is shown, and which action levels could be established for patient-specific quality assurance based on MGDR method. Materials and Methods The manufacturer uses a 10x10cm 2 beam to determine the density of PMMA and absolute calibration of the ArcCHECK and also to validate 3DVH®. This field is impossible in tomotherapy and our TPS doesn’t calculate plans with a static table and static gantry. Therefore, the following plans were planned (Tomotherapy v 5.1.1.6) on an ArcCHECK + Multiplug MVCT study: • Density determination: rotational plan for cylinder 16cm diameter and 10cm long (2Gy/fraction; jaw- 2.5cm; pitch-0.287; modulation factor: 1.8), including the central input diodes. This plan was recalculated in the DQA station module with forced densities ranging from 1.12 to 1.22 g/cm 3 . The planned dose versus the measured dose in the two central input diodes and the 1%/2mm absolute-local gamma of the dose distributions were compared. • PDP-algorithm acceptance ( Planned Dose Perturbation ): 3DCRT plan for cylinder 4cm diameter and 10cm long (1Gy/fraction; jaw-5cm; pitch-0.172) plus three rectangles directionally blocked. Planned vs. reconstructed profiles (AC-PDP virtual gel) and D exit /D entry ratio were compared. In addition, the algorithm was tested for geometries from TG119. Absolute calibration: cross-calibration with IC ExRadinA1SL and Virtual Water (depth 3.3cm) for a static beam of 40x5cm 2 and 30s of irradiation time. Retrospectively, ∆D (%) = D 3DVH – D TPS / D TPS ∙ 100, for D 98 , D 50 , D 1 , and ∆V(%) D = V 3DVH (%)- V TPS (%) for PTVs and OARs was evaluated in 23 patients (9 H&N; 14 SBRT). Results Density 1.15 g/cm 3 provided the best 1%/2m gamma index, 94.9%, and the lowest dose deviation (<0.3%). The small difference between D exit /D entry planned and calculated, 0.231 and 0.220, respectively, as well as an average difference Δ (D IC -D TPS ) of 0.23% ± 0.01%, for the 23 cases analyzed, confirm the assigned density. The profiles calculated in 3DVH agree with planned, fig. 1b, except for the maximum, due to the inherent ripple of tomotherapy plans, and the penumbra (Fig. 1a). The program allows to adjust the penumbra by a correction factor that did not improve the H&N result of the TG119 (fig. 1c). The PDP model would have to be adapted to our unit to get a good matching between the profiles and a D exit /D entry = 0.27–0.28 (according to manufacturer for 6MV), but it can’t be do it by the user.