ESTRO 38 Abstract book

S999 ESTRO 38

EP-1841 Couch modeling optimization for TomoTherapy planning and delivery M. Tanooka 1 , W. Okada 1 , K. Sano 2 , S. Mayuri 2 , H. Doi 3 , M. Miyazaki 1 , R. Nakahara 1 , M. Sueoka 1 , H. Suzuki 1 , M. Fujiwara 1 , N. Kamikonya 1 , Y. Inomata 2 , K. Yamakado 1 1 Hyogo College of Medicine, Department of Radiology, Nishinomiya, Japan ; 2 Takarazuka City Hospital, Department of radiotherapy, Takarazuka, Japan ; 3 Kindai University Faculty of Medicine, Department of Radiation Oncology, Sayama, Japan Purpose or Objective Carbon-fiber flat top couches have been widely used for radiotherapy. It was reported that the beam intensity attenuation inside the couch should be taken into account in treatment-planning system. From the start, TomoTherapy planning system (Accuray Precision) has been provided such a capability. However, this couch model does not have appropriate physical densities. Furthermore, it is not possible to correct these physical densities. Therefore, we performed a novel couch modeling optimization for TomoTherapy planning and delivery. Material and Methods The beam intensity attenuation caused by the TomoTherapy Radixact couch unit was evaluated by creating and adding a new optimized couch model in a planning support system (MIM Maestro). The couch modeling started with acquiring CT images, the images were transferred to the planning support system for contouring. Then these DICOM images data and a created DICOM RT structure set for couch modeling were transferred to the TomoTherapy planning system. Additionally, dose was compared between calculation and measurement in order to optimize the appropriate densities for the contoured couch region. To do this, a Exradin A17 chamber was placed on the couch. Doses were measured at gantry angles ranging from 0° to 355° in 5° increments for a 6 MV photon beam. This setup was replicated in the TomoTherapy planning system and dose was calculated using collapsed-cone-convolution (CCC) algorithm (TomoDirec plan, Forward Planning, dynamic jaw mode, Field size: 10×5 cm 2 ). Interactive appropriate physical density optimization was performed by comparing calculated and measured doses in order to minimize the discrepancies. Finally, the optimized appropriate physical densities were employed in the typical QA plan of 14 patients to evaluate the impact of the treatment couch on the dosimetry validation. Results The presence of the treatment couch in a plan resulted in a decrease ranging from 3.0% up to 11.1%, depending on the beam angles. The conventional modeling decreased at isocenter dose discrepancy in the range of 0.9% to 2.3%. However, our modeling decreased dose discrepancy less than 0.5% except for gantry angles 115° and 245° (Figure 1). Our simple shape model with optimized physical densities was selected among the models considered, and the optimized physical densities of the treatment couch were 1.3 g/cm 3 and 2.5 g/cm 3 . A total average decrease in absolute dose at a selected point of 0.2% was calculated in the QA plans. Additionally, the dose distribution our modeling yielded an average pass rate using gamma index (3%, 2 mm) of 96.5% (max 99.7%, min 91.7%), and the pass rate was higher than conventional modeling (Table 1). Figure 1

Table 1

Conclusion We could successfully obtain an accurate couch model for TomoTherapy planning system by interactively optimizing the physical density of the couch using planning support system. Furthermore, this modeling proved to be an efficient way of correcting the dosimetric effects of the treatment couch in TomoTherapy planning and delivery. EP-1842 Automated treatment planning of prostate cancer using prioritized clinical-goal based optimization R. Kierkels 1 , A. Fredriksson 2 , E. Hynning 2 , S. Both 1 , J.A. Langendijk 1 , B. Vanhauten 1 , E.W. Korevaar 1 1 University Medical Center Groningen, Radiation Oncology, Groningen, The Netherlands ; 2 RaySearch Laboratories, Stockholm, Sweden Purpose or Objective To increase departmental efficiency and achieve high quality treatment plans for prostate cancer patients. Therefore, we implemented fully automated prioritized clinical-goal based treatment planning (ATP) for intensity modulated radiotherapy (IMRT) and volumetric modulated arc therapy (VMAT) and evaluated the ATP plans with the clinical dosimetrist-optimized treatment plans. Material and Methods The clinical goals of prostate radiotherapy (patients containing gold markers) were extracted from our medical protocol and ordered by priority into a wishlist. The order and priorities of the clinical goals were fine-tuned based on the dosimetrist-optimized clinical VMAT plans of two previously treated prostate cancer patients. The final version of the wishlist was used by the ATP solution (RayStation v6.1, RaySearch Laboratories AB, Stockholm, Sweden) to automatically optimize treatment plans for novel patients. The ATP optimization is a lexicographic method that tries to fulfill the goals of the wishlist iteratively. A goal of higher priority will not be compromised to fulfill goals of lower priority. To assess the quality of the wishlist, prostate IMRT and VMAT plans were automatically optimized by the ATP algorithm for eight previously treated prostate cancer patients. From these patients a dosimetrist-optimized prostate VMAT plan was also available. The treatment planning time of a representative dosimetrist-optimized prostate case was approximately 180 minutes. Target coverage, conformity index (CI95%) and homogeneity index (HI98%) as well as organ at risk dose values were compared between the ATP plans and the dosimetrist- optimized plans.