ESTRO 2021 Abstract Book

S1451

ESTRO 2021

1 Radiotherapiegroep, Medical Physics, Deventer, The Netherlands; 2 Radiotherapiegroep, Medical Physics, Arnhem, The Netherlands

Purpose or Objective To develop and implement an automated, completely DICOM based process for the transfer of absolute couch coordinates and treatment site specific tolerance tables from the TPS RayStation to the R&V system Mosaiq. This replaces manual entering the tolerance table and makes overriding the couch interlock and couch copy for the first fraction unnecessary. Consequently, the workload for the RTTs was reduced and patient safety increased. Materials and Methods Our center treats 5000 patients annually using ten Elekta linacs and two Philips Big Bore CT scanners. They are all equipped with Long Extension Universal Couchtops (Civco) or iBEAM one couchtops (Medical Intelligence) with a 7 cm index and fixation system. During treatment patient immobilization devices are positioned using lock bars on fixed index locations as determined during acquisition of the planning CT. The absolute couch position with respect to the treatment isocenter is calculated. The couch height results from the vertical distance of the isocenter in the RTplan to the CT couch surface, whereas the lateral position results from the distance to the lateral CT couch center. The longitudinal couch position requires a one-time cross calibration of the CT couch position against the treatment couch position registered by the R&V system. The longitudinal CT couch position is available in a private DICOM tag of the CT image. This calibration has proven to be stable over time, but is unique for each CT simulator. The DICOM location tag is used to distinguish between them. The treatment couch position and the site specific tolerance table were inserted in the DICOM RTplan while it was transferred from the TPS to the R&V system using Python scripts.

The calculated and actual couch positions during treatment, as registered by the R&V system, were stored in a dedicated database and used to monitor both setup accuracy and the data processing pipeline. Results With the introduction of this automated process 5000 manual couch overrides and copy actions in our R&V system were made redundant per year. The R&V system activates an interlock when the patient is positioned outside the site specific tolerance distances compared to the calculated treatment couch position. Fig. 2 shows the differences between the actual and calculated treatment couch positions for 1514 treatment fractions of 131 breast cancer patients. For 93% of the fractions the setup deviation was within 1.5 cm lateral and longitudinal, and within 1.0 cm vertical. For the other 7% the RTTs had to give a conscious interlock override.

Conclusion We have successfully implemented an automated method to transfer couch coordinates and treatment site specific tolerance tables from the TPS to the R&V system. This process minimizes manual overrides in the R&V system and improves patient safety. Analysis of the differences between predicted and measured couch positions provides insight in patient setup accuracy and guidance for tolerance table selection. PO-1727 Biological validation using tissue samples of three CT scanners commissioned for proton therapy J.D. Azcona 1 , B. Aguilar 1 , A. Viñals 1 , P. Cabello 1 , J.M. Delgado 1 1 Clínica Universidad de Navarra, Servicio de Radiofísica y Protección Radiológica, Madrid, Spain Purpose or Objective To verify the accuracy of a CT stoichiometric calibration for proton therapy by comparing predicted SPR values with measured ones for a variety of biological tissue samples. Materials and Methods Three Siemens CT scanners (Somatom Drive, Emotion and Biograph PET/CT) were calibrated using the