Abstract book - ESTRO meets Asia

S5 ESTRO meets Asia 2018

and eyes were better with DSC 0.92(0.03) and 0.88(0.04) as compared to parotids 0.78(0.07).The accuracy of propagated contours is limited for complex deformations that include large volume and shape change, hence, visual validation of propagated contours is recommended for clinical implementation. Conclusion The DIR algorithm was commissioned and validated for adaptive contouring using AAPM TG132 protocol. VMAT Planning with Functional Lung Imaging N. Bucknell 1 , N. Hardcastle 2 , S. Prasad 1 , D. Ball 1 , T. Kron 2 , S. Siva 1 1 Peter MacCallum Cancer Center, Radiation Oncology, Melbourne, Australia 2 Peter MacCallum Cancer Center, Physical Sciences, Melbourne, Australia Purpose or Objective Functional lung imaging has the ability to enhance understanding of underlying lung function and facilitate tailoring treatment planning to an individual’s own lung function and potentially minimise toxicity. This planning study was conducted to evaluate if incorporating functional lung information could improve functional dose-metrics while boosting dose to the primary tumour and avoiding excess dose to organs at risk (OARs). This is the first reported planning study investigating minimisation of dose to both perfused and ventilated lung. Material and Methods 11 patients with locally advanced lung cancer with pre- treatment V/Q-PET/CT as part of a prospective observational clinical trial were planned using a VMAT technique in Eclipse 13.6 planning system (Varian Medical Systems, USA).(1, 2). High functioning (HF) lung was defined as the intersection of the top 70% of ventilated and perfused lung and functional (F) lung as the intersection of the remaining ventilated and perfused anatomical lung. 60Gy in 30 fractions was prescribed to the primary and nodal PTV and 69 Gy in 30 fractions to the primary tumour. Anatomical plans were optimised to anatomical lung and functional plans to the HF and F lung volumes. Heart volume receiving 50 Gy was limited to <25% and mean heart dose <20Gy.(3, 4) Proximal bronchial tree was limited to 1cc <64.5Gy. Other OAR constraints were consistent with our published protocol and RTOG guidelines.(2,5) Two tailed paired t-tests were used to test for significant differences between functionally adapted and anatomical plans. Results There were no significant differences in target coverage (PTV or primary tumour), mean heart and oesophagus dose, or near maximum heart, oesophageal and spinal canal doses. The only OAR constraint with a significant difference between plans was heart V50 with a mean difference of 0.11%; the highest heart V50 was 7.63% Both functional and to a lesser extent, anatomical lung doses were significantly reduced in functional plans. HF lung fMLD was reduced by a mean of 1.1Gy (p=0.002) and fV20 by 4.1% (p=0.02). F lung fMLD was reduced by a mean of 0.7Gy (p=0.0001) and fV20 by 2.3% (p=0.004). Table 1 summarises the statistically significant results showing a difference between anatomical and functional lung plans OC-017 Personalising Lung Radiation Therapy: Optimising

Conclusion Jaw tracking resulted in decreased dose to critical structures in IMRT and VMAT plans. But significant dose reductions were observed for critical structure in the JT- IMRT compared to SJ-IMRT technique. In JT-VMAT plans dose reduction to the critical structure were not significant compared to the JT-IMRT due to relatively lesser monitor units. OC-016 Commissioning and validation of deformable image registration (DIR) software using AAPM TG132 R. Phurailatpam 1 , K. Joshi 1 , D. Deshpande 2 , J. Sv 1 1 ACTREC-Tata Memorial Centre, Radiation Oncology Dept, Navi Mumbai, India 2 Tata Memorial Hospital-TMC, Medical Physics, Mumbai, India Purpose or Objective To commission and validate deformable image registration software (Velocity 3.2.0) using TG132 Material and Methods DIR (Velocity 3.2.0) was validated using TG132 protocol, using physical phantoms, virtual phantoms and clinical images. Solid Phantoms, made in-house with various known deformations and a set of 10 virtual phantoms were used. Deformation field vectors in lateral, AP and SI direction were evaluated for various organs and compared with the ground truth. Two clinical sites HN (n=9) and prostate (n=16) were validated. .Organs were manually delineated by a radiation oncologist, compared with the DIR generated contours. Dice Similarity Co-efficient (DSC) and Hausdorff distances Hf avg . were analysed. Results Mean (SD) DSC and Hf avg (mm) of the physical phantoms were 0.86(0.07) and 0.9(0.43)mm respectively. The virtual phantom resulted in consistent results for all the ROIs investigated. In prostate, the contours propagated for bladder resulted in better DSC of 0.91(0.04) as compared to rectum 0.85(0.05). However, in extreme cases when the change in bladder volume was more than 200 cc, DSC was 0.46. Similarly, in rectum, due to the presence of gas and faecal filling; DSCs were observed as 0.66 and 0.46 respectively. In HN, contours propagated for Brainstem