TU-C-BRE-06: Effect of Implementing In-House Treatment Couch Model On Patient Specific QA for Pinnacle SmartArc Treatment Plans




Failure to model the treatment couch during VMAT QA planar dose calculation may Result in discrepancies between measured and calculated dose. These discrepancies are due to beam attenuation by the treatment couch that is not included in dose calculation. This work evaluates effects of accounting for this attenuation on patient specific VMAT QA results using an in-house created Varian Exact couch model in Pinnacl


Patient specific VMAT QA results for 13 Pinnacle SmartArc plans generated for treatment on a Varian iX accelerator were studied. These plans included 3 treatment sites (7 H' N, 5 brain, 1 prostate). A Pinnacle model for Varian Exact couch was created in-house to replace the CT simulator couch. Composite arc planar doses were calculated with no couch present (NC) and with the Exact couch model (CM) in place for each plan. QA measurements were taken using IBA Matrixx Evolution ion chamber array set up in IBA MultiCube and were compared to each planar dose. Gamma passing criteria of both 3%/3mm and 2%/2mm tolerances were used.


Over all treatment sites, increases in gamma passing rates from NC to CM ranged from -0.4% to +27.3% at 3%/3mm and +0.1% to +30.5% at 2%/2mm. Mean increases in passing rates were +3.7% and +5.3% for 3%/3mm and 2%/2mm tolerances, respectively. Site-specific mean increases (NC to CM) in gamma passing rates were +4.4%, +3.4%, +0.4% (3%/3mm tolerance) and +6.9%, +3.7%, and +2.9% at (2%/2mm tolerance) for H' N, brain, and prostate, respectively.


Results support use of a couch model when generating planar dose for patient specific VMAT QA analysis. The improvements were most noticeable at 2%/2mm tolerance and for the H' N and brain sites. Eliminating treatment couch beam attenuation as a source of discrepancy in QA measurements may improve the ability to recognize otherwise masked delivered dose errors.