Encrypted login | home

Program Information

Investigation of VMAT QA Discrepancies Observed with a Deterministic Dose Calculation Algorithm and Commercial Diode Array

J Faught

J Faught*, Y Vinogradskiy , D Thomas , M Miften , D Westerly , University of Colorado Denver, Aurora, CO


TH-AB-FS1-10 (Thursday, August 3, 2017) 7:30 AM - 9:30 AM Room: Four Seasons 1

Purpose: To investigate clinical VMAT QA failures observed using a Boltzmann deterministic dose algorithm and a commercial diode array.

Methods: Twenty patients planned using Varian Eclipse and treated on a TrueBeam STx were selected for this study. Plans comprised two groups: those with maximum field sizes ≥10x10cm2 (N=10) and ≤5x5cm2 (N=10). Plans spanned a range of treatment sites and modulation factors, and were re-calculated using both 6 and 10MV beams. The AcurosXB algorithm was used to compute QA plans on the SunNuclear ArcCheck and ScandiDos Delta4 Phantom+ diode arrays with calculation grids of both 2mm and 1mm. Plans were delivered to both phantoms. Gamma analysis was performed using a 3%/3mm criteria and threshold of 10%.

Results: Modulation factor was not a significant predictor of gamma passing rate for 6 or 10MV plans. No significant differences were observed between 6 and 10MV passing rates. Measurements showed significantly lower pass rates for small vs. large field plans calculated with AcurosXB and delivered to the ArcCheck phantom (median: 90.4% vs. 99.2%, p<0.001). We did not observe lower passing rates for small fields using Delta4. Calculation of QA plans using 1 mm dose resolution or AAA algorithm resulted in improved passing rates for small field plans.

Conclusion: VMAT QA performed using AcurosXB and ArcCheck showed significantly lower passing rates for small vs. large field plans. Lower passing rates appear to be due to down-sampling of the transport grid by the AcurosXB dose engine to reduce calculation times. The down-sampling results in reduced accuracy away from the iso-center where the ArcCheck diodes are located and can explain poor passing rates with ArcCheck that are not observed with Delta4. While ArcCheck pass rates can be improved by using finer dose resolution or AAA, caution should be exercised when using measurements off-isocenter to verify AcurosXB small fields.

Contact Email: