ESTRO 2020 Abstract book

S740 ESTRO 2020

PO-1312 Diffusion-weighted imaging evaluation using a 3D-printed phantom for Stereotactic Radiosurgery T. Torfeh 1 , R. Hammoud 1 , Y. Arunachalam 1 , S. Aouadi 1 , N. Al-Hammadi 1 1 National Center for Cancer Care & Research, Radiation Oncology, Doha, Qatar Purpose or Objective Diffusion weighted (DW) MRI is increasingly being used for monitoring tumor response to radiotherapy and for tumor diagnosis. Within the context of radiosurgical treatment with CyberKnife® System (CK), MRI is used for tumor delineation which is a crucial moment in planning process and remains challenging due mainly to artifacts present in the DW images. The purpose of this work is to evaluate the geometric distortion of DW images using in-house 3D printed phantom. Material and Methods Glass tubes of 5 mm diameters positioned on 3D printed PLA base was used in this study. A structure composed of 12 holes in the superior inferior direction and 8 holes in the left right direction was 3D printed as shown in figure 1. Glass tubes filled with water were inserted inside the holes. The phantom covers a field of view of 300*150*150 mm 3 . DW images were acquired on a GE 1.5T MR-SIM unit. Single- shot echo-planar imaging with the following parameters (b=1000 s/mm 2 ; FOV, 240 mm; imaging matrix, 256×256; NEX=1, TR = 8000 and TE = 91.1 ms) was used. Images were acquired in the axial and sagittal planes. DWI was tested with respect to all diffusion directions and 2 directions of the frequency encoding gradient. Figure 1. A and b) the 3D printed phantom inside the MR unit and c) Diffusion weighted image. In-house software that automatically acquires measurements was developed. The software derives a distortion map of the MR DWI sequence by calculating the one-to-one correspondence between the locations of the tubes in the MR and their locations obtained from the CT. Results Our preliminary results showed that the mean geometric distortions are 1.1 mm. Our results showed no significant differences in the distortion when the acquisition plane or the diffusion direction are modified. Finally, higher distortion was observed when the frequency encoding gradient was along anterior/posterior direction. Conclusion The present work is a preliminary study aimed at designing a phantom for the assessment of the geometric distortion as part of a routine quality assurance program for DW-MR images. Further investigations on the influence of imaging parameters such as, TE, TR, b value… on the image quality will be carried out. Finally, the phantom will be used to allow the assessment of the DW with different fluids. PO-1313 Population-based virtual phantoms for paediatric radiotherapy applications – initial developments C. Veiga 1 , S. Moinuddin 2 , J. Gains 3 1 University College London, Centre for Medical Image Computing, London, United Kingdom ; 2 University College London Hospital, Radiotherapy Department, London,

In this study, the MCD method was applied to measure small field output factors for a stereotactic radiosurgery unit, a Leksell GK Icon. The measured 8 mm collimator relative output factor shows variations within +/- 3% across the MCD and SCD radiochromic film dosimetry methods. PO-1311 Dosimetric evaluation of a 2D multidetector dedicated to stereotactic radiotherapy R. Villeneuve 1 , T. Cardon 1 , O. Allegrini 1 , N. Garnier 1 , R. Amblard 1 , B. Serrano 1 1 Centre Hospitalier Princesse Grace, Division de Radiophysique et Radioprotection, Monaco, France Purpose or Objective The SRS MapCHECK (SRSMC - SunNuclear) coupled with the phantom StereoPHAN is a new multi detector dedicated to the control of SRS/SBRT plans delivered in radiotherapy. It is composed of 1013 silicon diodes. We propose to evaluate its dosimetric performances before focusing on clinical plan controls. Material and Methods Energies used were X6, X6FFF (Free Flattening Filter) and X10FFF. The pulse rates used were respectively 600, 1600 and 2400MU/min and the UM number was 100. Calibration accuracy was studied by comparing 10 measurements with SRSMC oriented at 0° and 180°. The repeatability of the central diode was evaluated by calculating the Relative Standard Deviation (RSD) of 20 consecutive readings. The dose linearity response was determinated by measuring the response for a number of monitor units (MU) ranging from 10 to 10 000. The pulse rate dependence was evaluated by measuring the response for different repetition rates ranging from 10 to 600MU/min for X6, 80 to 1400MU/min for X6FFF and 80 to 2400MU/min for X10FFF. The response for dose per pulse was evaluated by varying source-detector-distance from 70cm to 130cm and comparing the results with an ionisation chamber (0.3cc PTW). The angular response was evaluated with the SRSMC inserted in the StereoPHAN and for all around the phantom. The field size dependence was evaluated for field of 1 to 7cm² and compared with 5 averaged GafChromic films XD (Aschland) measurements. The dose distributions of 110 RapidArc treatment plans calculated with the TPS Eclipse (Varian) were measured and compared. Results Concerning calibration accuracy, the diode difference reached 0.7% for X6, 1.2% for X6FFF, 1.3% for X10FFF and the maximum difference value for 99% of diodes was respectively of 0.41%, 0.76% and 0.9%. For repeatability the RSD was of 0.05% for all energies and the linearity RSD was of 0.33% for X6, 0.23% for X6FFF and 0.24% for X10FFF. For pulse rate response, maximum difference was respectively of 0.6%, 1% and 1.6% for a dose rate of 80MU/min. With a very low dose rate of 10MU/min the difference was of 3.6% for X6. For dose per pulse response the difference between SRSMC and 0.3cc reached 0.8% for X6, 0.3% for X6FFF and 0.6% for X10FFF. Concerning angular dependence, for all energies, the deviation was less than 1% excepted for angles 85° to 100° and 260° to 280° for which a difference of 2.3% was obtained. Regarding field size response, SRSMC and films results agree to less than 1.6% for X6, 0.6% for X6FFF. For X10FFF, the deviation is under 2% exepted for 1x1cm² size for which we reached 2.7%. For plan comparisons, the rate of plans meeting the criteria (95% of the points and a threshold of 10%) were 98.2% (3%/2mm), 90% (3%/1mm) and 80% (2%/2mm). Conclusion SRSMC is an interesting 2D detector solution, user-friendly comparing to gafchromic films, for the control of stereotactic radiotherapy treatment plans. However, it has some limitations for very small field sizes, very low dose rate and for beam angles around 90° and 270°.