ESTRO 38 Abstract book

S90 ESTRO 38

train a support vector regression model that predicts DVH PCs from OVH PCs (see Figure 1). In our clinical workflow the planQC tool is initiated when the dicom data of a final clinical plan is exported from the treatment planning system to the planQC DICOM node, where the planQC tool application is executed. The application calculates the predicted and planned DVH curves and generates a personalized scorecard in a PDF report. The report is added to the departmental oncology information system (OIS) and reviewed by the planner. Threshold levels of 3 Gy for dose and 3% for dose and volume metrics at the OARs are set to detect if a plan is eligible for replanning. Plans are replanned by changing the treatment technique template for each of the OAR dose parameters that can be improved and starting the Auto-Planning again.

Figure 1

Results The planQC tool was incorporated into the clinical workflow of automated prostate planning on March 1st 2018. Currently, 40 clinical prostate plans have been checked by planQC. For 8 of the 40 patients the plans did not pass the check and further optimization of the plan was done. In Figure 2 an example of a patient plan is shown where the planQC tool detected a suboptimal mean dose for the bladder (upper panel). After replanning the mean dose of the bladder was reduced by 5.3 Gy (lower panel), from 28.2 Gy to 22.9 Gy. In general, the mean improvement for the 8 plans was 3.6 Gy/% for the metrics of the OARs which were improved.

Conclusion iRex significantly established dose-response relationships with grade 2-4 late GI and/or GU toxicities in retrospective data, especially iRex60 and iRex70. Regarding of pelvic organ geometrical uncertainties, with further investigation, the proposed Vex, Rex, and iRex concept could probably be used as the novel IGBT dose constraints in addition to D2cc, D1cc, and D0.1cc.

Proffered Papers: PH 3: Proffered paper: New developments in automated treatment planning

OC-0179 Clinical implementation of plan quality control for automated prostate planning M. Kusters 1 , F.J. Dankers 1 , P. Van Kollenburg 1 , R.J. Smeenk 1 , R. Monshouwer 1 1 Radboud University Medical Center, Academic Department of Radiation Oncology, Nijmegen, The Netherlands Purpose or Objective Currently most of our radiotherapy plans are automatically generated with a single treatment technique template in the Auto-Planning module in Pinnacle 16.0.2 (Philips Healthcare, Fitchburg, WI, USA). An inhouse-developed plan quality control tool (planQC) was built to check the quality of our automated prostate plans. This tool predicts a personalized dose-volume histogram (DVH) for each organ-at-risk (OAR) based on the anatomy of each individual patient. In this study the planQC tool was clinically implemented. Material and Methods Historical data of 129 automatically planned clinical prostate patient plans which all fulfill the clinical criteria were used to train and test a prediction model. In all these plans one single VMAT arc was used (95 to 265º), by which 70 Gy is given in 28 fraction of 2.5 Gy. The model was trained on 100 plans and validated on a separate test set of 29 plans based on a random split. The model was made by first calculating principal components (PCs) of the DVH and overlapping volume histogram (OVH), where the latter is a measure to capture individual patient anatomical information. Second, the calculated PCs were used to

Figure 2:Result before (upper) and after replanning (lower panel) Conclusion A planQC tool was successfully introduced in the clinical workflow and monitor the quality of automated prostate plans. It can be used to detect plans that deviate from model development cohort and can therefore possibly be improved. For 8 of the 40 patients the plan was labelled suboptimal and could be improved.