ESTRO 2021 Abstract Book

S1555

ESTRO 2021

Conclusion A fast generalized intelligent system has been proposed for segmenting bone marrow in CT volumes using a multilabel continuous convex relaxation approach. The preliminary results obtained show that the proposed algorithm outperforms some state-of-the art methods. The algorithm will help physicians and physicists in the radiation planning and the analysis of potential constraints associated with haematological toxicity. PO-1825 Plan quality versus anatomic mutual disposition in moderate hypofractionated prostate radiotherapy S. Trivellato 1 , P. Caricato 1,2 , V. Faccenda 1,2 , G. Montanari 1 , V. Tremolada 1 , R. Lucchini 3,4 , D. Panizza 1,4 , S. Arcangeli 3,4 , E. De Ponti 1,4 1 ASST Monza, Medical Physics Department, Monza, Italy; 2 University of Milan, Department of Physics, Milano, Italy; 3 ASST Monza, Department of Radiation Oncology, Monza, Italy; 4 University of Milan Bicocca, School of Medicine and Surgery, Monza, Italy Purpose or Objective Moderate hypofractionated radiotherapy can be considered as a standard treatment for low and intermediate-risk prostate cancer. To evaluate the treatment plan’s quality, retrospectively selected plans have been analyzed to find relationships between the obtained dose-volume values and the mutual disposition of critical organs at risk (OARs) and the target. Materials and Methods Fifty prostate cancer patients were treated between October 2019 and October 2020 with an Elekta VersaHD linear accelerator delivering 60 Gy in 20 fractions. The planning target volume (PTV) was created by a uniform 7 mm-expansion of the prostate except for the reduced posterior margin of 5 mm. Other contoured OARs are bladder, rectum, bowel, and femoral heads. Monte Carlo treatment plans were optimized by 4 medical physicists with the multicriterial optimization of Elekta Monaco TPS (v5.10) defining mathematical functions and relative weights for a 6 MV 330°-double arc. The process is repeated until an acceptable target coverage is reached and OAR dose constraints are fulfilled. An expansion-intersection volume (EIV) for rectum and bladder has been defined as the intersection volume between the target and the OAR isotropically expanded by 5 mm (rectum EIV and bladder EIV). The relationship of EIVs with the corresponding obtained dose-volume values has been evaluated to highlight the 3D geometric relationship between contoured structures. Results The median PTV was 142 cm 3 [76-146] with a median value of mean dose of 60.7 Gy [59.9-61.6], and median V 95% = 99.0% [90.1-100.0]. The median rectum and bladder volumes were 50 cm 3 and 140 cm 3 , respectively. The dosimetric review have showed an acceptable linear regression between the rectum V 46 Gy and the rectum EIV with R 2 =0.47 (Figure 1). No significant regression has been found for the rectum V 37 Gy . The bladder dose-volume values have not showed a good correlation with the bladder EIV, but with the bladder volume and with the bladder minus the bladder EIV volume: V 48 Gy and V 41 Gy vs bladder volume R 2 =0.47 and R 2 =0.48, respectively; V 48 Gy and V 41 Gy vs bladder minus bladder EIV R 2 =0.53 and R 2 =0.54, respectively (Figure 2).