Click here to


Are you sure ?

Yes, do it No, cancel

The Impact and Detectability of MLC Positioning Error in the VMAT Plan Delivered On Varian Halcyon

S Gay*, T Netherton , C Cardenas , P Balter , L Court , UT MD Anderson Cancer Center, Houston, TX


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

Room: Exhibit Hall | Forum 7

Purpose: To assess (1) the impact of MLC positioning errors on dose distributions for the new Halcyon MLC design, and (2) our ability to detect these errors with portal dosimetry.

Methods: Random and systematic MLC position errors were introduced into 21 lung VMAT plans. To add random errors, each MLC control position was changed by a uniform random value between ±3, ±5, ±7, or ±10mm. To add systematic errors, 3, 5, 7, or 10mm symmetric shifts in the same direction were made to all MLC positions in the plan. The plans were recalculated with constant MU, and the dose distributions were compared. Portal dosimetry was then performed for each plan (189 total QA plans) using the Halcyon system, and the delivered fluence was compared against the predicted fluence of the original treatment plan with a gamma index of 3%/3mm.

Results: PTV coverage (D95) was reduced by less than 3% for 100% of plans for which the random error was 7mm or less, and for 90% of plans for which the systematic error was 3mm. Mean lung dose was rarely affected by random or systematic errors, with only 3% of plans having greater than 10% change. The median portal dosimetry pass rate for the original plans was 99.1% (s.d:2.5%, range:90.0–99.85%). This reduced with increasing MLC error by an average of 2%/mm for random MLC errors, and 6%/mm for systematic errors. The percentage of plans passing a 95% gamma criteria was 95% for the original plans, 67% and 47% with 3mm and 5mm random errors, respectively, and 5% and 0% with 3mm and 5mm systematic error, respectively. Plan complexity or number of arcs did not significantly affect pass-rate.

Conclusion: Portal dosimetry on the Halcyon detects MLC position errors before they are large enough to impact patient dosimetry (for lung patients).


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email