ESTRO 2020 Abstract Book

S180 ESTRO 2020

Conclusion Difference in marker position between pre-treatment CBCT and post-treatment CBCT varied substantially between patients. Marker motion of up to 1.0 cm in CC direction was seen during a single DIBH despite use of narrow external gating window and visual feedback. OC-0341 Seminal vesicle motion tracking in 3D cine-MR during MR-Linac prostate treatments D. De Muinck Keizer 1 , T. Willigenburg 1 , M.D. Den Hartogh 1 , J.R.N. Van der Voort van Zyp 1 , B.W. Raaymakers 1 , J.J.W. Lagendijk 1 , H.C.J. De Boer 1 1 UMC Utrecht, Radiotherapy, Utrecht, The Netherlands Purpose or Objective We develop MR-guided stereotactic ablative prostate radiotherapy (SABR) with on-line adaptation of intrafraction motion based on real-time 3D cine-MR imaging. Previously, we presented a method for accurate automatic prostate tracking based on soft tissue contrast (ESTRO 38). Now we present a method for soft tissue contrast based tracking of the seminal vesicles from 3D cine-MR dynamics. To our knowledge, this is the first study that quantifies full 3D seminal vesicle intrafraction motion. Material and Methods Nine patients with low-intermediate risk who underwent adaptive prostate RT on a 1.5 T MR linac (Unity, Elekta) had 3D cine-MR imaging during dose delivery of each fraction. Each cine-MR dynamic was acquired with a balanced 3D gradient echo sequence with a resolution acquisition time between 8.5-9.37 seconds per 3D frame. The daily delineation of the first acquired scan was used as a mask for the prostate in the cine-MR imaging sessions, while additional individual (left and right) seminal vesicle masks were created on the first cine-MR dynamic as these were not part of the clinical CTV. Next, an in-house developed algorithm performed soft tissue tracking of the prostate and seminal vesicles in subsequent cine-MR dynamics using a mutual information metric and rigid transformations of each mask separately. Results The algorithm was applied to 1861 dynamics over 36 fractions acquired during beam on, with a mean duration of 7:58 minutes. Significant vesicle intrafraction motion was observed in anterior-posterior (AP) and cranial-caudal (CC) translation directions with significant rotations about the left-right (LR) axis as presented in Fig. 1. Seminal vesicle translations showed larger spread than prostate corpus for all time points in both AP and CC directions (Fig. 1). Especially LR rotations of seminal vesicles were significantly increased with respect to the prostate rotations with mean difference over 3 degrees and 95% spread ranging over 15 degrees. An overview of the intrafraction motion visualization tool is provided in Fig. 2, showing the prostate and vesicle masks, rotation, translation and full 3D visualization of both seminal vesicles and prostate. Conclusion We have developed a tracking algorithm for seminal vesicles in cine-MR data acquired during MR-guided prostate RT (MRgRT). The presented method provides quantitative intrafraction motion data of seminal vesicles that may feed subsequent plan adaptation. This can lead to margin reduction for the seminal vesicles in prostate SABR of high risk patients.

treatment CBCT was 0.3 cm (range 0.3 to 1.4 cm). Stability examination on pre-treatment repeat DIBH CTs was not sufficient to guarantee per-treatment stability cf. patient 3 and 4 in Figure 1. The mean time between the pre-treatment CBCT and post-treatment CBCT was 14 min (range 8 to 20 min). For planar kV images it was not possible to evaluate marker position in one patient due to lack of image contrast. The maximum difference in marker position (CC direction) on planar kV images during one fraction was between 0.8 to 1.3 cm for the four patients and maximum difference was between 0.3 to 1.0 cm within one DIBH (see Figure 2). The markers moved cranially within each DIBH in three out of four patients.