ESTRO 2024 - Abstract Book

S3348

Physics - Detectors, dose measurement and phantoms

ESTRO 2024

A highly modulated VMAT clinical plan (spine SBRT) was chosen. The plan consisted of 2 6X-FFF full arcs. Dose calculation was performed on a CT image data set from an anthropomorphic male Alderson-RANDO phantom using ECLIPSE TPS (Varian vs 15.6.05) Acuros-XB Dm,m. A Varian TrueBeam with a Millenium 120-MLC collimator and equipped with AS1200-EPID (electronic portal imaging device) was used for treatment delivery.

Image 1. On the left, anthropomorphic male Alderson-RANDO phantom, molded of tissue-equivalent material and designed to accurately simulate patient treatment conditions. On the right, clinical spine SBRT case recalculated on RANDO’s TC. PerFraction (SunNuclear) transit IVD system was used. The integrated images from EPID were automatically retrieved by PerFraction which compares these transit 2D dose images with the predicted dose image (projection of the planned beams through the RANDO planning CT at the plane of EPID) in a forward projection approach. To assess the sensitivity of PerFraction system in detecting failures, for the chosen plan, changes simulating treatment delivery errors were intentionally introduced. In particular we forced errors due to the machine and errors due to the patient. For machine errors we used the ones suggested by Lehmann on their SEAFARER study. DICOM headings of the plans were modified so that PerFraction could link these plans (with machine-related errors) to the original plan without errors. For patient errors we focused on patient’s setup and increase of patient diameter by adding bolus. To simulate the positioning errors, shifts of 3, 5 and 7 milimetres were applied in all directions: vertical, longitudinal and lateral. To simulate anatomical changes, different layers of bolus (1 and 2 centimetres) were added over the area to be treated on the phantom. The original plan, without errors, was exported to PerFraction. Pre-treatment verifications were performed for all the plans (original and with errors) and compared with the original plan predicted dose. Subsequently, RANDO phantom was placed on the treatment couch. Positioning was verified using CBCT. The original plan and the plans with machine related errors were delivered. Finally, we treated RANDO with the original plan, but introducing positioning errors and increasing the diameter by adding bolus. EPID was set at 60 cm from the isocentre and an integrated image was collected for each treatment arc. The metric we used for the 2D IVD analysis was a Local Gamma evaluation (3%, 3mm, 30%Th), gamma passing rate of 95%. For pre-treatment verifications we used a Local Gamma evaluation (3%, 2mm, 20%Th), gamma passing rate of 95%.