ESTRO 36 Abstract Book

S775 ESTRO 36 _______________________________________________________________________________________________

Results Across all treatment sites, the mean gamma index was 99.6% for the calculated dose and 98.3% for measured dose. For each of the treatment sites evaluated, the computed dose typically showed closer agreement with the Eclipse TPS calculation than the measured dose. This study demonstrated that for the Prostate and Node treatment site the average difference in gamma index between the computed and measured dose was within - 0.51%. This was -1.22% and -2.02% for Head and neck and Brain treatment sites respectively. Conclusion This result verified that the IBA Compass system is sufficiently accurate and has been adopted for RapidArc treatment plan verification based on either measurements, computation or both. EP-1452 Evaluation of a collapsed-cone algorithm in a commercial software for in vivo volumetric dosimetry J. Gimeno Olmos 1 , V. Carmona 1 , F. Lliso 1 , B. Ibanez- Rosello 1 , J. Bautista 1 , J. Bonaque 1 , J. Perez-Calatayud 1 1 Hospital Universitari i Politecnic la Fe, Radiotherapy department, Valencia, Spain Purpose or Objective Dosimetry Check (DC) (Math Resolutions) commercial software performs pre-treatment and transit EPID-based dosimetry. It provides a verification of treatments, being of interest due to the benefits of the in vivo volumetric dosimetry, which guarantee treatment delivery and anatomy constancy. In this study, the performance of a newly introduced collapsed-cone (CC) dose calculation algorithm is evaluated, as compared with the currently available pencil beam (PB) algorithm and with a conventional Treatment Planning System (TPS) and ionisation chamber measurements. Material and Methods The commercial version of DC (v.4.11) is only CE and FDA cleared for PB algorithm. The CC algorithm is being used as a beta version (v.5.1). To test if the CC algorithm considers heterogeneities correcty, measurements were done in the IMRT Thorax Phantom (CIRS), which simulates a human thorax. It has several inserts for ionisation chamber measurements. Six plans were generated, similar to the already published work for the PB commissioning (Phys Med 30: 954-9). Three with the isocentre in the phantom centre (isocentre A, tissue equivalent): (1) four open 10x10 cm static fields in box configuration, (2) 10x10 cm rotational field, (3) typical lung clinical treatment (patient A); and three centred in the phantom’s left lung (isocentre B): (4) and (5) equivalent to (1) and (2), (6) typical lung clinical treatment (patient B). The plans were delivered in a Clinac iX (Varian) accelerator equipped with EPID aS1000, acquiring cine images, which were then converted to fluence by DC to finally calculate dose with PB and CC algorithms. The plans were also calculated in the TPS Eclipse v.13.0 (Varian) with AAA and Acuros XB algorithms. Calculated point doses were compared against ionisation chamber measurements, performed in the isocentre for each plan with a PinPoint chamber model 31006 (PTW). DC dose distributions were also evaluated against TPS (Acuros algoritm) using 3D gamma analysis (3% global/3 mm) for the structure defined by the 95% isodose. Results Results are shown in table 1. As expected, CC algorithm improves PB results, mainly in isocentre B where the heterogeneities have greater effects. For isocentre A, the mean difference improves from 0.6% for PB to -0.2% for CC, while for isocentre B, it improves from 6.5% to -0.8%. A very significant improvement in the gamma analysis is

also observed. Figure 1 shows an example of dose distribution. It has to be mentioned that the calculation time for CC algorithm is of the order of hours, making this algorithm not yet suitable for routine patient verifications. An improvement is expected by the manufacturer to allow GPU calculations.

. Conclusion The possibility of in vivo evaluation and the potentiality of this new system have a very positive impact on improving patient QA. CC algorithm provides much better results in heterogeneous cases, but it is at the cost of a higher computation time. Improvements are also required in the integration of DC with the R&V system. EP-1453 Modeling a carbon fiber couch in a commercial Treatment Planning System R. Gómez Pardos 1 , D. Navarro Jiménez 1 , A. Ramírez Muñoz 1 , E. Ambroa Rey 1 , M. Colomer Truyols 1 1 Consorci Sanitari de Terrassa CST, Radiotherapy, Terrassa, Spain Purpose or Objective With the increased use of carbon fiber couch tops and the raise of techniques like VMAT with considerable dose delivered from posterior angles, currently their modeling is strongly recommended (Report of AAPM Task Group 176). The main objective of this work is to model the iBEAM® evo Couchtop in the TPS Monaco. The second goal is to assess the overall impact of not using the couch in VMAT calculations comparing gamma passing rates with an Octavius4D phantom (PTW, Freiburg, Germany). Material and Methods The modeling was made for an Elekta Synergy LINAC with Agility head equipped with the iBEAM couch. The EasyCube homogeneous phantom (Euromechanics Medical Gmbh, Nuremberg, Germany) was placed centered on the couch and aligned with the isocentre. The charge was measured with a Farmer ionization chamber every 5 gantry degrees, 100 MU/field, 10x10 cm 2 field size, for both 6 and 15 MV. All the measurements were corrected by pressure and temperature. The relative to zero gantry degree attenuation was calculated for every gantry angle analyzed. Previously the absolute dose at 0 gantry angle was measured. The couch was modeled by an outer shell of carbon fiber (CF) and an inner part of foam (Foam). The same measured fields were calculated in the Monaco TPS with the Monte Carlo algorithm, 1% Statistical Uncertainty per Control Point, 2 mm grid spacing, dose to water. Then the