ESTRO 37 Abstract book

S100

ESTRO 37

(Modus QA, Canada), driving a sinusoidal motion with 15 mm aplitude and 10s period. Motion was applied in parallel with the leaves’ motion axis. On continuously acquired EPID images (approx. 10Hz), positions of both ball bearing and tracked circular aperture were detected. Sinusoidal fits yielded the an estimated tracking latency of the entire tracking system, including MR-imaging, image processing and MLC motion. Results

Figure 1 shows the excellent object visibility in both MRI and EPID, which enabling the system latency determination. Tracking experiments with different MRI- frequencies (Figure 2) showed the strong dependency of the system latency on MRI-sampling frequency (92ms for 8Hz, 213 ms for 4Hz). Due to coarse temporal sampling and spurious MRI signals, we observed oscillatory tracking In this work we successfully demonstrate the first real- time MRI-guided MLC-tracking on the 1.5T MR-linac system. The performance of the tracking system was determined and validated using the independent portal imager. Ongoing work focuses on optimizing the tracking latency with respect to MRI-sequence types. Furthermore, fast and more stable image processing are under investigation, in order to suppress oscillatory effects and improve system latency. responses. Conclusion OC-0190 Surface guided radiation therapy for breast cancer improves accuracy without the need for skin marks C. Russell 1 , H. Mack 1 , S. Paul 1 , S. Senthi 1 1 The Alfred, Alfred Health Radiation Oncology, Melbourne- Victoria, Australia Purpose or Objective Patient positioning for breast radiation therapy has traditionally relied upon permanent or temporary skin marks. However, permanent tattoos increase the incidence of body image dissatisfaction and psychological distress following treatment, while temporary pen marks may fade and be inaccurately redrawn. Surface guided radiation therapy (SGRT) using an optical system can provide a 3D coordinate reconstruction for patient setup without requiring skin marks, and has the additional benefit of intra-fraction monitoring. We compared accuracy of SGRT setup with setup utilising skin marks. Material and Methods Patients undergoing radiation therapy following breast conservation and not needing nodal irradiation or deep inspiration breath hold technique were included in this retrospective study. Prior to acquiring the AlignRT SGRT system (VisionRT, London, UK), patients were set up using temporary skin marks (Cohort 1). Following its acquisition, patients were set up using temporary skin marks for initial gross positioning, followed by SGRT Proffered Papers: RTT 2: Reproducibility optimisation in clinical practice

OC-0189 First MLC-tracking on the 1.5T MR-linac system M. Glitzner 1 , P.L. Woodhead 1 , J.J.W. Lagendijk 1 , B.W. Raaymakers 1 1 UMC Utrecht, Department of Radiotherapy, Utrecht, The Netherlands Purpose or Objective MRI-guided radiotherapy promises better treatment options due to the greatly improved observation capabilities of MRI as compared to other on-line modalities. In addition to fine contrast in anatomy and pathology, MRI offers relatively fast imaging frequencies, enabling real-time feedback control of the delivery beam for motion-compensated treatments. Currently, MRI- guided treatment devices support respiratory gating, triggered by the anatomy as seen in the MR-images. In this work, we present first results of MRI-guided MLC- tracking using a clinical 1.5T MR-linac sytem. We determine tracking latency and investigate its sensitivity to MRI sampling schemes and control settings. Material and Methods All experiments were performed on the clinical prototype of the Elekta Unity MR-linac (Elekta AB, Sweden). The machine features a 1.5T high-field MRI and a 7MV linac. The beam is shaped using an MLC with 80 leaf pairs moving in fixed y-direction (IEC 1217). For all experiments, a circular aperture with 40 mm diameter was applied. Every 40 ms, leaf position updates were sent to a prototype linac control system using an in-house developed tracking software. In parallel to steering the leaves, the tracking software processes a real-time stream of images from the MRI- machine with 4 Hz and 8 Hz, respectively. For each image, the position of a moving object was determined using a center of gravity algorithm. Accordingly, the MLC was steered to reach the new position. The object consisted of cast agarose gel and a circonium oxide ball bearing immersed in the gel and was thus fully visible in both MRI and EPID. The non-metallic bearing guaranteed minimal MR-image distortions. The object was moved using a QUASAR MRI4D motion phantom