ESTRO 2020 Abstract book

S743 ESTRO 2020

Excellent reproducibility was determined for the experimental setup. The correction factors determined can be used for dosimetric measurements on MR-linacs as well as for benchmarking new Monte Carlo methods. The results show the importance of the exact determination of magnetic field correction factors for correct use in clinics. PO-1318 Impact of Document Scanner Resolution on Raw Signal Noise in Radiochromic Film Dosimetry S. Devic 1 , N. Tomic 2 , S. Aldelaijan 2 , J. Seuntjens 2 , B. Moftah 3 1 Jewish General Hospital, Oncology, Montreal, Canada ; 2 McGill University, Oncology, Montreal, Canada ; 3 King Faisal Specialist Hospital and Research Centre, Department of Biomedical Physics, Riyadh, Saudi Arabia Purpose or Objective Although widely acceptedfor the film to be the highest resolution 2D dosimeter not much work have been done so far to investigate limits of spatial resolution of this dosimetry system. One question usually asked is what scanning resolution one should use. Implicitly, one would assume answer to this question would depend on deemed measurement resolution in a given experiment, commonly defined by the size of the region of interest (ROI) over which an average pixel value is calculated. In this work, we attempt to answer the question what would be the optimal scanning resolution that minimizes raw pixel value noise for high resolution dose measurements. Material and Methods Three pieces of EBT3 model GafChromic TM films, irradiated to doses of 0.2, 1, and 10 Gy, were scanned with document scanner using transmission mode with 127, 254, and 508 dpi scanning resolutions. Two square ROIs (0.5 mm and 1 mm in size) were sampled (according to Fig.1.a) and all 3 color channels have been extracted for analysis. For each ROI, different scanning resolutions will result in different number of pixels, in accordance to table given in Fig.1.b. It is important to emphasize that for a given ROI size pixel values are always sampled over the exact same region on the film since film pieces were not moved between. Over each ROI we calculated the average value (m) and standard deviation (sd), and then noise using expression: Noise [%] = 100 x (sd/m).

kBMQ. Monte Carlo simulations of this factor show deviations from experimentally determined values of more than 1 % (van Asselen et al. 2018). The aim of this work was the experimental determination of kBQM factors for magnetic flux densities between 0 T and 1.5 T for different A large electromagnet (Bruker ER073W) was installed in front of an Elekta Precise linear accelerator (6 MV). The response of different detectors (PTW 60019, PTW 31021, PTW 31010, PTW 30013, PTW 34001) was measured inside a water phantom that was placed between the pole shoes of the magnet at a 10 cm water-equiavalent depth. The radiation field axis, the symmetry axis of the ionization chambers and the direction of the magnetic field were perpendicular to each other in pairs. The magnetic field strength was varied in steps of 0.15 T up to a magnetic flux density of ±1.5 T. Each measurement was repeated three times Results The experimental results are shown in fig. 1. The results show a large qualitative difference between the investigated chamber types. For smaller sensitive volumes, a reduced response can be observed, while detectors with larger sensitive volumes show an increased response. In addition, the change in response is strongly dependent on the orientation of the magnetic field, especially for ionization chambers with a small sensitive volume. The magnetic flux densities for which the measured signal changes most or least are shown with the corresponding values for kBMQ in table1. ionization chambers. Material and Methods

Figure 1: Experimental values for kBQM. Negative magnetic flux densities represent a deflection of secondary electrons to the chamber tip, positive magnetic flux densities to the chamber stem.

Results Figures 1.c and 1.d show noise of raw transmission signal (pixel values) at three dose levels obtained from scanned images using three scanning resolutions for square ROIs with sizes of 0.5 and 1 mm respectively. Different colors correspond to different color channels in scanned images. It is apparent that under all conditions, noise of the raw signal increases with dose, which is expected since the transmission signal decreases when dose increases. For larger measurement resolution of 1 mm ROI size (Fig.1.d), the lowest noise at all dose levels and for all color channels is observed for 127 dpi scanning resolution, which

Table 1: kBQM values for the magnetic flux densities for which the measured signal changes most (maximum deviation from 1) and for the magnetic flux densities for which the measured signal changes least (minimum deviation from 1). Conclusion