ESTRO 38 Abstract book

S583 ESTRO 38

0.58±0.49 mm was detected. Mean deviations to the reference plan of 0.48±0.31 mm and 0.58±0.35 mm were found for the first and second observer, respectively. The length of the catheter traces varied 0.45±0.39 mm on average. The mean deviation of CI varied by 0.25± 0.29%, variations in DNR, COIN, mean heart dose and mean lung dose were all <0.01 %. The skin dose changed in the maximum by 4.10 %. Nevertheless, manual reconstruction of the catheter paths is a time consuming (3-5 minutes per catheter) procedure in treatment planning. Evaluation of magnitude and implication on the geometrical deviations and the dose distribution using the EMT measurements is still ongoing.

higher dose to the sigmoid, as also demonstrated by the wider error bars of the sigmoid parameters plotted in figure. In the opposite, the V150% to the target resulted lower with ACE, probably in light of the higher attenuation of the applicator material if compared to water. Differences between uniform densities and HU-based calculations were all negligible, as well as dose distributions resulting with the use of MVAs of two different diameters.

Conclusion In general, on the 24 investigated treatments the ACE dose calculation model results in similar dose distributions as the commonly used AAPM-TG43 algorithm, both for targets and organs at risk, with the exception of the V150% to the target and the dose at 5 mm above the applicator tip. PO-1048 Variability in catheter reconstruction for multi-catheter interstitial brachytherapy K. Kallis 1 , T. Kaltsas 1 , S. Kreppner 1 , M. Lotter 1 , V. Strnad 1 , R. Fietkau 1 , C. Bert 1 1 University Hospital Erlangen, Radiation Oncology, Erlangen, Germany Purpose or Objective Accelerated partial breast irradiation of the tumor bed using interstitial brachytherapy is a common treatment option for selected breast cancer patients. Reconstruction of the source path is a time consuming manual procedure during treatment planning that influences the planned dose delivery. The aim of this study is to evaluate inter- observer variabilities for catheter reconstruction. Further, we explored the feasibility of an automated catheter reconstruction using an electromagnetic tracking (EMT) system, integrated into an afterloader prototype. Material and Methods The catheter reconstructions of three physicists were compared for 26 patients, in total 426 catheters, see figure 1. Keeping constant the CT data, contours, dwell times, dwell positions (DPs), plan optimization and normalization settings, the geometrical deviations between the corresponding DPs of the repeated reconstruction and the reference plan were evaluated. Also, the effects on coverage index (CI), dose non- uniformity ratio (DNR), conformal index (COIN), and the dose to the organs at risk were analyzed. The clinical approved plan served as reference. In addition, the catheter traces were reconstructed with the help of EMT data. The EMT measurement was conducted directly after CT imaging using a Flexitron prototype (Elekta, Veenendaal, The Netherlands) equipped with an EMT sensor. In combination with an Aurora EMT system (NDI, Waterloo, Canada) all catheters can be automatically digitized, which takes 6-9 minutes depending on the number of implanted catheters. Using an in-house routine, the measured catheter traces were rigidly registered to the catheter button centers defined on the CT image and automatically imported to the treatment planning system within one minute. Results Over all patients and all manual catheter reconstructions a mean deviation between corresponding DPs of

Conclusion The study proved that a repeated reconstruction of the catheter traces does not lead to a large change in dose exposure or large geometrical deviations. However, lack of ground truth poses a challenge to reliable quality assessment of the reconstruction. Using EMT measurements seems to be a feasible, dose free and quick option for catheter reconstruction. PO-1049 Assessing PTV margin adequacy in permanent breast seed implant for complex target geometries A. Guebert 1,2 , A. Frederick 1,2 , M. Roumeliotis 1,2,3 , T. Meyer 1,2,3 , S. Quirk 1,2,3 1 University of Calgary, Department of Physics and Astronomy, Calgary, Canada ; 2 Tom Baker Cancer Centre, Division of Medical Physics, Calgary, Canada ; 3 University of Calgary, Department of Oncology, Calgary, Canada Purpose or Objective Permanent breast seed implant (PBSI) brachytherapy presents a convenient, single day alternative to standard whole breast radiotherapy. For widespread adoption, robust treatment planning recommendations are required. Clinical PBSI treatment plans recommend a 10-15 mm expansion of the clinical target volume (CTV) to the planning target volume (PTV). The PTV margin is cropped to the chest wall muscle and 5 mm from the skin surface as these structures present anatomic boundaries. For patients where the target is located in close proximity to the skin and chest wall, significant cropping of the PTV is required on both sides. For this subset of patients, accounting for approximately 25% of our cohort, poor coverage has been observed on the post-implant CT. This simulation study aims to investigate PTV cropping in the skin-chest wall direction and the impact on post-implant dosimetry. Material and Methods Four typical CTV volumes (6.6, 10.1, 15.6, and 30.5 cm 3 ) were modelled as ellipsoids. The PTV margin for each volume was modelled as isotropic expansions of 10 mm in the direction parallel to the needle, and the template left- right direction, and 0, 4, 7, or 10 mm in the skin-chest wall direction (Figure 1). Treatment plans were generated with 15 mm seed spacing for needles passing through the CTV, and 10 mm seed spacing for needles passing through only the PTV. The prescription dose was 90 Gy. For each margin combination, 1500 implants were simulated by application of a seed delivery error sampled from our reported clinical