ESTRO 2020 Abstract book

S863 ESTRO 2020

The coverage of the primary and nodal target volumes was comparable for both techniques and for both subsets of patients. The primary planning target volume (PTV) receiving at least 95% of the prescription isodose ranged from 97.2±1.1% to 99.1±1.2% for H and from 96.5±1.9% to 98.3±0.9% for HT. For the nodal CTVs the dose received by 98% of the planning target volume ranged 55.5±0.1 to 56.0±0.8Gy for H and HT respectively. The only significant and potentially relevant differences were observed for the bowels. In this case V 40Gy resulted 226.3±35.9 and 186.9±115.9 cm 3 for the node positive and node negative patients respectively for Halcyon. The corresponding findings for HT were: 258.9±60.5 and 224.9±102.2cm 3 . On the contrary, V 15Gy resulted 1279.7±296.5 and 1557.2±359.9 cm 3 for HT and H respectively for node positive and 1010.8±3209 vs. 1203.8±332.8cm 3 Conclusion This study confirmed the equivalence between Halcyon based and Helical Tomotherapy based plans for the intensity modulated rotational treatment of cervix uteri cancer. Different levels of sparing were observed for the bowels with H better protecting in the high-dose region and HT in the mid-low dose regions. The clinical impact of these differences should be further addressed. PO-1508 Exploration of deformable image registration to augment training data for deep learning contouring R. Baggs 1 , P. Aljabar 1 , M. Gooding 1 , P. Poortmans 2 , Y. Kirova 2 1 Mirada Medical Limited, Science, Oxford, United Kingdom ; 2 Institut Curie, Radiation Oncology, Paris, France Purpose or Objective This work investigates whether deformable image registration of CT images, as a means for data augmentation, affects deep learning networks for auto- contouring structures in breast cancer patients. Augmentation can be used to improve convolutional neural networks for image segmentation, for example in [1], which deployed model-based augmentation for MRI, rather than CT, in a non-radiation therapy (RT) application. Material and Methods Simulation CT images were acquired for 50 patients undergoing RT for breast cancer. Manual contours of the breast and lymph node regions (axilla levels 1 to 4 and interpectoral) served as ground truth labels for evaluation. A Deep Learning Contouring (DLC) model was trained on 40 cases, with the remaining 10 split equally into cross- validation (CV) and test sets. The model was trained using jittering [2], i.e. applying random affine 2D transformations to training CT slices. During training, each batch of slices was augmented with the same number of jittered slices. The same 50 cases were then used to create a larger training set using offline deformable registration based on optical flow [Mirada RTx, Mirada Medical Ltd]. 451 registrations were randomly selected from the 2450 combinations, excluding the original CV/test data and gross registration errors. Incorporating the original 50, gave a total of 501 datasets. This final augmented dataset was used to train another DLC model, evaluating against the original CV and test sets to enable comparison. The accuracy of each method was measured using quantitative measures, the Dice similarity coefficient (DSC) and the median surface distance (SD). Results The DSC results showed that for each organ, the data augmentation method using deformable registration was comparable or better than the standard jittering method. Surface distances were comparable, or lower, for deformable registration showing a slight improvement in accuracy when using this method. The jittering method under-segmented the breast in two test cases, as seen in the figure’s outliers.

Figure 1: A model of the ideal prostate treatment plan. A KB of 562 clinical prostate VMAT treatment plans (generated and QA-ed in 2017 and 2018 according to a well-established protocols) were used to inform this model. The mean and 25 th percentile of the PTV-OAR dose gradients from the training dataset were used to generate two sets of optimisation objectives in Pinnacle and these were tested on a cohort of 142 patients from 2019. The new treatment plans for 10 of the test patients (for which all of prostate radiotherapy dose criteria were met) were reviewed visually by a medical physics expert experienced in prostate treatment planning. Results For 114 of the 142 test cases, at least one of the new plans (average or 25 th percentile) met all of the dose criteria for prostate radiotherapy at The Christie. Furthermore, for 105 of the test cases, at least one of the new plans was superior to the original clinical plan. In Figure 2, PTV- Rectum dose gradient is plotted against PTV-Bladder dose gradient for the KB training data (transparent red), clinical test data (solid red), average new treatment plans (yellow) and 25 th percentile new treatment plans (blue). Figure 2: Treatment plans generated using KB dose gradient optimisation (blue and yellow) show more consistent and steeper PTV-OAR dose gradients than corresponding clinical treatment plans (red). Visual review of the new treatment plans for 10 patients from the test cohort showed that all were considered clinically acceptable. Conclusion A novel approach to KB treatment planning for prostate VMAT has been proposed, trained and tested. Initial results show that treatment plans generated automatically from the predictions of the model are clinically acceptable in more than 80 % of cases and show superior and more consistent OAR sparing than their corresponding clinical treatment plans. Although there is scope for refinement of the predictions, this method shows promise for realising significant efficiencies within treatment planning departments and with the transfer of knowledge between centres. PO-1507 VMAT by means of a dual layer stacked multileaf collimator or helical tomotherapy for cervix uteri L. Cozzi 1 , P. Subhajit 2 , S. Jamema 2 , C. Supriya 2 , M. Arjun 2 , F. Antonella 1 , A. Jai Prakash 3 1 Humanitas, Radiotherapy, Rozzano, Italy ; 2 Advanced Centre for Treatment Research and Education in Cancer- Tata Memorial Centre, Radiation Oncology, Mumbai, India ; 3 Advanced Centre for Treatment Research and Education in Cancer- Tata Memorial Centre, Radiation Oncology, Mumbai, Italy Purpose or Objective To ascertain the dosimetric performance of a new delivery system (the Halcyon system, H) equipped with dual layer stacked multileaf collimator (MLC) for risk adapted targets in cervix uteri cancer patients compared to another ring- based system in clinical operation (Helical Tomotherapy, HT) Material and Methods Twenty patients were retrospectively included in a treatment planning study (10 with positive lymph nodes and 10 without). The dose prescription (45 Gy to the primary tumor volume and a simultaneously integrated boost up to 55Gy for the positive patients) and the clinical planning objectives were defined consistently as recommended by an ongoing multicentric clinical trial. Halcyon plans were optimized for the volumetric modulated arc therapy. The plan comparison was performedby means of quantitative analysis of the dose volume histograms Results