ESTRO 38 Abstract book

S59 ESTRO 38

diagnostic challenge because both entities have similar imaging characteristics on conventional MRI. Functional imaging techniques, such as dynamic contrast enhanced (DCE) MRI or positron emission tomography (PET), could overcome this diagnostic dilemma. A third study will be presented to investigate the potential of DCE-MRI and PET in discriminating high-grade glioma from RN in rats. Induction of RN was achieved by irradiating the right frontal region with 60 Gy using multiple arcs. Results suggested that functional imaging can be used to discriminate glioblastoma (recurrence) from RN. SP-0111 On-line MRI-guidance for dose accumulation and plan adaptation Bas Raaymakers 1 1 University Medical Center Utrecht, Radiotherapy, Utrecht, The Netherlands Abstract text There are two commercial, clinical hybrid MRI radiotherapy systems operational. The promise of these systems is to improve on the targeting accuracy by providing soft-tissue contrast images directly from the treatment table. These MRI data can be acquired both prior to- and during beam delivery. Such on-line visualization enables image guided Radiotherapy (IGRT) as we know it, but then based on soft- tissue contrast images. This presentation will mainly focus on the next steps of using on-line MRI data. The MR images can be used to reconstruct the delivered dose by combining the MRI with the linac output. This can be done at various time scales. Daily MRI can be used for dose accumulation per day. Intra-fraction MRI, in combination with time-stamped linac output can yield time-resolved dose reconstruction at an intra-fraction level. The goal is to track the dose delivery in real-time, for this MRI, linac readings and dose calculations need to be optimized. The accumulated dose can be used for treatment response assessment but clearly also as a metric for plan adaptation. Plan adaptation based on daily MRI from the treatment table is a next step for IGRT, the plan is adapted to the patient instead of trying to re-position the patient to the pre-determined plan. Similar to dose accumulation, the frequency of adaptation can be increased by also using intra-fraction MRI. The anatomical changes have to be interpreted live, various options are under investigation to allow real-time digestion of MRI, including faster MRI acquisitions and hybrid patient modelling-MRI techniques. Plan adaptations can be triggered by anatomical changes, but also by accumulated dose metrics. A loop to use the real-time dose accumulation as input for continuous (intra-fraction) plan adaptation will be presented. In summary, this presentation aims to show that on-line MRI guidance as a next step for soft-tissue based IGRT is only the first step towards continuously adapted radiotherapy so moving targets can be treated as if they are static. SP-0112 First clinical experience and future directions of MR-guided radiation therapy D. Zips 1 1 University Hospital Tübingen, Radiation Oncology, Tübingen, Germany Abstract text Since September 2018 the 1.5 T MR-Linac system is in clinical use. I will discuss our first clinical experiences and reflect on expectations as well as on our long-term vision for MR-guided adaptive radiation oncology. In addtion, I will discuss the potential of MRgRT to facilitate transformation of current radiation oncology into RO4.0.

OC-0113 MRI artifact simulation for clinically relevant MR sequences for guidance of HDR brachytherapy E.Beld 1 , M.A. Moerland 1 , M.A. Viergever 2 , J.J.W. Lagendijk 1 , P.R. Seevinck 2 1 UMC Utrecht, Department of Radiotherapy, Utrecht, The Netherlands; 2 UMC Utrecht, Image Sciences Institute, Utrecht, The Netherlands Purpose or Objective The purpose was to investigate the potential of several clinically relevant MR sequences for application during the HDR brachytherapy intervention to facilitate tracking/localization of brachytherapy devices (HDR source/titanium needle), which could simultaneously be used to visualize the anatomy. Simulations of the MR artifacts were implemented for a spoiled gradient echo sequence, a spin echo sequence, a balanced steady-state free precession (bSSFP) sequence and a bSSFP sequence with SPAIR fat suppression. Material and Methods Simulations Simulation of the artifact for the spoiled gradient echo sequence was performed as described in[1]. To simulate the other sequences, the implemented MRI signal equation was adapted. For a spin echo sequence, no dephasing effect was included because of the 180º refocusing pulse. For bSSFP, the steady-state signal equation was included. For SPAIR, the spins at a frequency of the SPAIR pulse were set to 0. MRI acquisition The object, a non-active Ir Flexisource (Elekta)/a titanium needle (Ø1.9mm, Elekta), was positioned in the center of a doped water phantom. MR imaging was performed on a 1.5T MRI system (Ingenia, Philips). 4 types of 2D MR sequences were applied (2 intersecting 2D slices): spoiled gradient echo, spin echo, bSSFP and bSSFP-SPAIR (see Table1). Furthermore, the MR sequences of the clinical prostate HDR brachytherapy scan protocol were applied: a 3D bSSFP-SPAIR sequence, a T2-weighted and a T1- weighted spin echo sequence (both multi-slice 2D), see Table2. Angles of 0º and 20º between source/needle and The object position was determined by template matching between the MR images and the simulated artifacts, using a subpixel phase correlation algorithm[1]. A 2D phase correlation was performed for the 2D slices, and a 3D phase correlation for the 3D and multi-slice 2D acquired volumes. The average object position (of 7 obtained positions) and the mean deviation from this average were calculated. B0 were applied. Post-processing

Results Fig.1 shows the simulated and the MRI images for the 4 applied 2D MR sequences (with the object positions overlaid on the images). These results demonstrate that the simulations highly correspond to the acquired images