ESTRO 2020 Abstract Book

S181 ESTRO 2020

OC-0342 Experimental validation of real-time rotation- including dose reconstruction during tumor tracking C.G. Muurholm 1 , T. Ravkilde 2 , R. De Roover 3,4 , S. Skouboe 1 , R. Hansen 2 , W. Crijns 3,4 , P.J. Keall 5 , T. Depuydt 3,4 , P.R. Poulsen 1,6 1 Aarhus University Hospital, Department of Oncology, Aarhus, Denmark ; 2 Aarhus University Hospital, Department of Medical Physics, Aarhus, Denmark ; 3 KU Leuven, Department of Oncology, Leuven, Belgium ; 4 University Hospitals Leuven, Department of Radiation Oncology, Leuven, Belgium ; 5 The University of Sydney, ACRF Image X Institute, Sydney, Australia ; 6 Aarhus University Hospital, Danish Center for Particle Therapy, Aarhus, Denmark Purpose or Objective Hypofractionation in prostate radiotherapy increases the need for accurate treatment delivery, but motion can deteriorate the delivered dose. Real-time motion adaptation with MLC- or couch-tracking can reduce the dosimetric effect of motion by adjusting for translational target motion. The effects of residual uncompensated rotations are mostly unknown. In this study, we develop and demonstrate real-time rotation-including dose reconstructions performed in real-time during prostate radiotherapy phantom experiments. The real-time dose reconstructions are benchmarked against radiochromic film measurements of the same experiments. Material and Methods DoseTracker is an in-house developed program which uses a simple pencil beam algorithm to perform motion- including dose reconstruction during treatment delivery based on streamed tumor positions and linac parameters. DoseTracker’s ability to predict errors to a translating and rotating target was compared to film in three different scenarios: (1) No motion correction, and translational motion correction with (2) MLC tracking and (3) couch tracking. In each scenario, dose reconstruction was performed in real-time during delivery of two VMAT dual- arc plans, with 4mm CTV-PTV margin and intraprostatic tumor boosts. The plans were delivered to a CIRS pelvis phantom that reproduced three patient-measured motion traces. The motion was executed by pitch rotation of a drum with an embedded insert that contained 21 EBT3 film layers spaced 2.5mm apart. DoseTracker repeatedly calculated the actual motion-including dose and the planned static dose increment since the last calculation with a resolution of 1mm in all film planes. The experiments were performed with a TrueBeam accelerator with MLC- and couch-tracking based on the Calypso system. The motion-induced errors were quantified by comparison with static doses using the 2D 3%/2mm γ failure rate with a 10% dose cut-off as well as the motion- induced reduction in D95% (ΔD95%) for the CTV and GTV and the motion induced increase in urethra D0.1cc (ΔD0.1cc). Results Online real-time dose reconstruction was done in more than 80,000 calculation points at a rate of about 5Hz. Fig 1A shows dose color washes from one of the plans for both film and DoseTracker. Fig 1B illustrates DoseTracker’s ability to do time-resolved dose-reconstruction by showing the cumulative dose as a function of time to a point on the edge of the CTV. Fig 2 compares DoseTracker dose reconstruction and film measurements for the motion- induced dose error metrics used. The root-mean-square error of DoseTracker relative to film was 3.6% (γ failure rate), 0.12Gy (CTV ΔD95%), 0.28Gy (GTV ΔD95%), and 0.25Gy (urethra ΔD0.1cc).

Fig. 1: Most relevant intrafraction motion results for prostate and vesicles (left) with vesicle differences to prostate (right). Error bars provide the 95% spread.

Fig. 2: Adapted capture from the intrafraction motion (IM) tool, in which the IM from the prostate corpus and seminal vesicles is shown in 3D.