Click here to


Are you sure ?

Yes, do it No, cancel

Using a QA Test System to Determine Mechanical Tolerance Limits of Linacs for Highly Complex Treatment Plans

J Stroom*, A Taborda , S Vieira , C Greco , Champalimaud Centre for the Unknown, Lisbon, Lisbon


(Tuesday, 7/31/2018) 9:30 AM - 10:00 AM

Room: Exhibit Hall | Forum 3

Purpose: Treatment plans are still getting more modulated and complex. In this study, we used a QA test system to determine new mechanical tolerance limits of linacs that treat these plans.

Methods: Software was developed that converts dicom treatment plans to XML format that our Varian TrueBeam linac can read. The software can add acquisition of dosimetry images and introduce plan errors (ER) for sensitivity studies of QA systems. Multiple plans can be irradiated using one XML file to increase measurement efficiency. We used 22 IMRT/VMAT plans with varying prescriptions, energy and field sizes. We introduced systematic MLC errors (MS) from -2(close) to +2(open)mm, random MLC errors (MR) from 0 to 5mm (1SD), collimator errors (CO) from -6 to 6 degrees, and MU errors from -6% to +6%. Each plan was measured using the Varian pre-treatment portal dosimetry with 2%/2mm gamma analysis, yielding mean gamma (MG) and gamma pass rates (GPR) per beam. Error sensitivity was approximated by straight fits MG/ER and GPR/ER. We then determined those errors that yielded the same gamma changes as a 2% output error (i.e. current tolerance limit).

Results: Average MG and GPR for the original plans was 0.25 and 97.5%, respectively. Error sensitivity varied for different sources, yielding average values for GPR/CO=6.1%/deg, GPR/MR=1.0%/mm, GPR/MS=24.2%/mm, and GPR/MU=2.4%/%. Hence, the same GPR-drop is observed for e.g. MU=1% and MS=0.1mm. MG results were similar. Assuming the 2% MU error as reference, new tolerance limits are for CO 0.8deg, MR 4.7mm, MS 0.2mm. The MS-limit is much smaller than currently applied (1mm), probably due to the ever-smaller treatment segments.

Conclusion: Our QA test system showed that especially the MLC tolerance limits should be stricter for high-tech treatments. We will apply the system to other detectors and validate findings by studying the dosimetric effect of errors in the TPS.

Funding Support, Disclosures, and Conflict of Interest: This research was partly funded by Varian


Quality Assurance, Portal Imaging


TH- External beam- photons: Quality Assurance - IMRT

Contact Email