ESTRO 2021 Abstract Book

S1472

ESTRO 2021

Purpose or Objective The Surface Guided RadioTherapy (SGRT) allows monitoring patient position during treatment, as a complement to IGRT. Stereotactic treatments require submillimeter accuracy. The purpose of this work is to describe the quality assurance (QA) program designed for a SGRT system to be used during cranial stereotactic treatments, as well as its results. Materials and Methods The QA program has been implemented since February 2020, for a AlignRT (Vision RT) SGRT system installed on a TrueBeam (Varian) linac. The SGRT system used consists of three stereoscopic positioning optical devices (POD). Daily, the consistency of the position detected by each POD is checked to be <0.5 mm, using a 2D pattern of black circles on a white background. Weekly, the accuracy of the EPID positioning is checked to be lower than 0.5 mm, using Varian’s IsoCal tool that analyzes the position of the radiation isocenter and imaging devices from multiple projections of a cylindrical phantom with bb o its surface. Subsequently, the AlignRT isocenter calibration is verified through the analysis of EPID images for AP/PA and RL/LR projections of a cubic phantom containing five radiopaque spheres of 7.5 mm diameter. The system is recalibrated if any DOF exceeds 0.1 mm or degree of rotation. Next, the AlignRT accuracy with isocentric couch rotation is verified, analyzing AP/PA EPID images for +45º and +90º couch positions. Results Daily QA data showed average deviation of 0.3 mm between PODs, without observing a trend. In three cases, the tolerance was exceeded making necessary to recalibrate AlignRT isocenter. IsoCal results showed maximum deviations of 0.1 and 0.33 mm for the EPID and the radiation isocenters, respectively. The results of the AlignRT isocenter verifications were < 0.1 mm and 0.1 º in 24% of the cases (fig. 1), not being necessary to recalibrate the system weekly. The corrections applied were < 0.5 mm, except on 7 occasions (18%) in which the displacements were about mm. These cases coincided with a previous optical calibration, performed with the 2D pattern of circles (see vertical lines in fig.1).

Fig. 1. AlignRT deviations without isocentric couch rotation. The discrepancies found between AlignRT and EPID images, with isocentric couch rotation, were lower than 0.5, 0.4 and 0.3 mm in 99%, 92%, and 65% of cases, respectively (see fig. 2). No systematic deviation was detected (median value of 0.2 mm), although maximum differences were observed for + 90º couch rotation.