ESTRO 38 Abstract book

S1169 ESTRO 38

M.A. Benito Bejarano 1 , F. Saez Beltran 1 1 Hospital Provincial de Zamora, Department of Radiation Physics, Zamora, Spain Purpose or Objective Modern external radiotherapy techniques rely on different computerised systems to design and deliver the treatments. Treatment plans are created in a Radiotherapy Planning System (RTPS), and upon approval they are sent to a Record and Verify System (R&V), where the plan data is stored for subsequent use in each treatment fraction. A quality assurance (QA) procedure must be applied to ensure that there have been no errors in the plan design, as well as inconsistencies or data corruption during the transmission or storage of the plan parameters. In the present, most QA procedures are based in manual and time-consuming verifications. We propose an automated system to check the plan integrity and validate its proper transmission to the R&V. Material and Methods The plans were created in XiO RTPS. They were transmitted by means of DICOM RT PLAN protocol to the Mosaiq R&V. A set of routines was developed in MATLAB. Those routines enable the extraction and analysis of data from DICOM files generated by the RTPS, as well from the OPENRTP text files exported by the R&V. The system works in a three stages sequence. In the first stage, the "verifydicomrt" routine analyses the DICOM file from the RTPS, to seek inconsistencies or plan errors. In the second stage, "dicomprocess" extracts data from the DICOM file to be used in our in-house secondary monitor unit (MU) verification spreadsheet. In the third stage, the "rtpdicomrtcomp" routine compares selected parameters from the RTPS-generated DICOM file with those in the OPENRTP file from the R&V. In any of those stages, when an issue is detected a warning message is displayed. DICOM and OPENRTP files with intentional errors were created to test the system.

Table 1: Parameters processed by each routine Conclusion The usage of this checking system enabled the automation of plan QA verification in a quick and reliable way. Errors could be detected during the treatment plan design, transmission and storage stages. The system could also detect data modifications in the R&V database. EP-2113 Congruence of mechanical, radiation, and imaging isocentres of two types of Elekta linacs C. Mekala 1 , A. Naga 1 , N. Babu 1 , S. Kumar 1 , N. Khater 1 , C. Birch 1 1 Southampton University Hospital NHS Trust, Radiotherapy Physics, Southampton, United Kingdom Purpose or Objective In this study we quantify the congruence of mechanical, radiation and imaging isocentres of two Elekta linear accelerator linacs as it is crucial for cranial hypo- fractionated stereotactic radiosurgery and radiotherapy deliveries where lesions of small volumes are treated with high-dose per fraction of 5 to 21 Gy. Material and Methods Winston Lutz test was performed as a part of stereotactic pre-treatment linac quality control checks using a metal ball bearing phantom of diameter 5 mm which was set on the couch using in-room lasers which were originally calibrated to the MV isocentre. 2D planar MV images were acquired at 4 cardinal gantry angles and 4 couch angles with gantry set to 0⁰ for a field size of 2.4x2.4 cm 2 on an Elekta Beam Modulator linac and 3x3 cm 2 using a recently commissioned 6MVFFF beam on Elekta Agility. The images were analysed using Pipspro software and the ball bearing was moved to match with the centre of the radiation isocentre based on optimal shifts calculated by the software. A cone beam computed tomography (CBCT) image and exactrac image of the ball bearing phantom were also acquired and the congruence of all the isocentres was evaluated based on the optimal 3D shifts of the centre of the ball bearing phantom to that of the MV isocentre. The data was acquired on both the linacs on 25 stereotactic treatment days over a period of six months. Results Maximum systematic total deviation of 1.2 mm (0.6 mm radius) and 1.5 mm (0.75 mm radius) was found in the in

Figure 1: Plan transmission and automated QA procedures workflow Results "verifydicomrt" was able to catch common planning errors in parameters related to the couch, collimator, gantry, MU, or others (see table 1). This allowed the early detection and correction of those errors before the plan was transmitted to the R&V. "dicomprocess" enabled the automatic extraction of the data needed for independent MU calculation, minimising the possibility of manual data input mistakes. "rtpdicomrtcomp" was able to compare the data generated by the RTPS with the actual parameters stored in the R&V database. This permitted tracking down any transmission errors, data corruption in the RTPS or R&V databases, or any mistakenly manual modification of parameters. "rtpdicomrtcomp" can be run in a weekly basis to keep track of unintended changes in the plans stored in the R&V database.