ESTRO 38 Abstract book

S1072 ESTRO 38

for each dataset. For each camera-object distance four repeat datasets were acquired to assess reproducibility. Further measurements were performed with the object moving on a Standard Imaging respiratory gating platform (Standard Imaging Inc, USA) with a range of 1D motion amplitudes (5, 20, and 40 mm) and frequencies (0.18, 0.22 and 0.29 cycles/sec). Results The results of the camera depth measurements on the static object are shown in table 1. For distances between 300 and 600 mm the camera was able to measure the distance to an accuracy better than 1 mm. The measured depth values over a 10 x 10 pixel region of interest on the object had a maximum variance of 5 mm (0.5%) at a camera-object distance of 1000 mm. The variance was less for shorter camera-object distances. Figure 1 shows the measured depth as a function of time for motion amplitudes of 5, 20, and 40 mm with a frequency of 0.29 cycles/sec. The measured amplitudes of the motion from the camera depth data in figure 1 were 4.2, 19.4 and 41.6 mm respectively.

are comparable to CBCT setup errors [1.4 +/-1mm vs 1.3 +/-0.09mm in vert (p= 0.87), 1.4 +/-1mm, vs 1.7 +/- 0.8mm in long (p= 0.75), 1.5 +/-1mm vs 1.6 +/- 1 mm in lat (p= 0.53) directions respectively]. Mean rotational errors documented by OSMS and CBCT system are [1.02 +/- 0.8° vs 1.2 +/-0.8° pitch (p= 0.12), 0.6 +/-0.4°, vs 0.7 +/- 0.6° roll (p= 0.77), 0.96 +/-0.5° vs 0.98 +/- 0.6° in rotation (p= 0.82) directions respectively]. Mean translational errors documented by post treatment OSMS are comparable to CBCT setup errors (0.3 +/- 0.4 mm vs 0.5 +/-0.4 mm in vert, 0.5 +/-0.2 mm vs 0.6 +/-0.5mm in long and 0.3+/- 0.4 mm vs 0.8 +/-0.3mm in lat direction respectively). Mean rotational errors post treatment documented with OSMS and CBCT (0.4 +/- 0.5° vs 0.3 +/- 0.4° pitch, 0.3 +/- 0.4° vs 0.3 +/- 0.3° in roll, 0.3 +/- 0.2° vs 0.2 +/- 0.3° in rotation respectively). Average intrafractional translational errors were (0.5+/- 0.3 mm in vert, 0.7 +/-0.2 mm in long and 0.5 +/- 0.4mm in lat directions respectively), and mean rotational errors (0.45 +/- 0.3°, in pitch, 0.3 +/- 0.3° in roll and 0.33 +/-0.2° in rotation respectively). Pearson’s correlation coefficient for pretreatment setup verification by OSMS compared to CBCT was (r = 0.12) in vert, (r = 0.16) in long, (r=0.24) in lat, (r= 0.5) in pitch, (r = 0.04) in roll and (r= -0.41) in rotation. Figure (1): Frame-less, open mask Stereotactic Radiosurgery system with OSMS Conclusion Setup accuracy of shifts computed by OSMS is comparable to CBCT data. Overall, there is positive correlation between OSMS and CBCT documented setup errors (0< ƍ <1). EP-1965 RealSenseTM Camera Technology for Real Time Surface Monitoring During Radiotherapy A. Fielding 1 , Y. Jonmohamadi 1 , S. Lee 1 , A. Pandey 1 1 Queensland University of Technology, Science and Engineering Faculty, Brisbane, Australia Purpose or Objective The aim of this work is to investigate the use of the latest generation of the Intel® Real SenseTM SR300 camera technology for real-time measurement of radiotherapy patient positioning. The systems make use of optical camera technology that measures the distance to a surface; rapid image acquisition can then enable real time motion detection of the location of the surface. The measurement based study investigated the accuracy and precision of the camera system through a series of controlled measurements of static and moving objects. Material and Methods This work is a feasibility study of the use of an Intel® Real SenseTM SR300 camera for measuring patient position and motion for radiotherapy applications. The camera was connected to the USB port of a laptop running the Windows 10 operating system. Frame capture was performed using the free RealSenseTM Software Development Kit (SDK 2.0) and cross-platform libraries. 30 frames were captured at a rate of 30 fps with each frame having 640x480 pixels. Measurements were performed on the flat surface of a static object with a range of camera- object distances between 100 and 1000 mm. The 30 frames were averaged and the mean and variance of the depth values in a 10 x 10 region of pixels was calculated

Conclusion The results of this preliminary work showed the Intel® RealSenseTM SR300 camera to be able to measure the respiratory like 1D cyclical motion of a test object with an amplitude of 5 mm with an accuracy better than a mm. The camera was able to operate at camera-object distances of up to 1000 mm with pixel noise levels of 0.5% making the camera a feasible option for measuring patient position and motion during radiotherapy. EP-1966 Deep inspiratory breath hold versus free breathing techniques in breast cancer radiotherapy A.D. Tawfik 1 , A. Ali 1 , S. Talima 1 , M. Mousa 1 1 Cairo University, Department of Clinical Oncology, Cairo, Egypt Purpose or Objective It has been theorized that radiation therapy to the breast using deep inspiratory breath hold (DIBH) technique can reduce cardiac doses by moving the heart away from the target volume. In addition, lung expansion during inspiration can decrease the relative volume of irradiated lung. To test this hypothesis, dosimetric and geometric comparisons were conducted between DIBH and conventional free breathing (FB) treatment delivery techniques in breast cancer patients.