Abstract Book

S160

ESTRO 37

parotid glands increased with decreasing planning priority from 30 ± 11 Gy (P0) (mean ± SD) to 38 ± 13 Gy (P14), which benefitted the oral cavity, larynx and MCI (Fig 1 top). Decreasing the planning priority of the parotids led to a large increase in prediction error (SD) of the corresponding KB models, from only 1.6 Gy in the absence of inter-OAR dependency (P0) to 5 Gy for low priority (P14) (Fig 1 bottom). Especially the oral cavity benefitted from lowering the parotid priority, expressed by both a decrease in Dmean from 60 ± 7 Gy (P0) to 48 ± 11 Gy (P14) (Fig 2 top), and a decrease in prediction error (SD) of the corresponding KB models from 5 Gy to 2 Gy for the (Fig 2 bottom).

inter-OAR dependency on the prediction accuracy of KB models will also propagate to the performance of KB automated planning. To improve KB planning, more sophisticated KB models are needed.

Proffered Papers: RTT 3: Imaging and protocols for treatment verification

OC-0306 Dose-recalculation for Head and Neck patients; a RTT's perspective S. Ali 1 , S. Beek van 1 , S.R. Kranen van 1 , J. Sonke 1 , P. Remeijer 1 1 Netherlands Cancer Institute, Radiotherapy, Amsterdam, The Netherlands Purpose or Objective Head and neck cancer (H&N) patients frequently show anatomical changes during the course of radiotherapy. Typically, these changes are detected in CBCT scans by visual inspection. The dosimetric consequences, however, are challenging to quantify as CBCT scans lack robust Hounsfield unit calibration. Therefore we developed a dose recalculation workflow, in which deformable image registration (DIR) is used to propagate the HU and contours of the planning CT to the geometry of the CBCT. This generates a virtual CT with corresponding structure set for dose-recalculation and evaluation in the treatment planning system (TPS). Since DIR accuracy may vary between patients, a visual inspection of the DIR by the RTT has been built into the workflow. In this study we assess the feasibility of visually inspecting the quality of the DIR in order to develop a DIR inspection protocol. Material and Methods 147 CBCT scans for 25 H&N cancer patients were retrospectively selected in consecutive order. The majority of the patients were treated in 35 fractions of 2 Gy with VMAT. CBCT-to-planning CT DIRs were performed with in-house developed software based on B-splines deformations and correlation ratio. The quality of the DIR was scored on three selected patient elements: 1) patients outer contour, 2) bony anatomy and 3) soft tissue, accumulating to a percentage of DIR-correctness: A (=not OK for dose-recalculation) and B (=OK for dose- recalculation, for this all the elements should be scored as correct). Figure1:

Results The DIR and generation of a virtual CT with corresponding structures took on average 10 minutes. The automatic dose recalculation in the TPS required another 15 minutes, after which it is available for evaluation. The scoring methodology was fast and easy to perform and may readily be expanded with further evaluation criteria. Of 147 DIRs, 94% were categorized as suitable for dose- recalculation, while 6% were not suitable to use for dose- recalculation or were needed density override of the

Conclusion The prediction errors of KB models are low in the absence of inter-OAR dependency (SD < 2 Gy), but strongly increase up to a SD of 5 Gy as the competition between sparing different OARs increases. The large influence of