Click here to


Are you sure ?

Yes, do it No, cancel

Independent Treatment Planning System Beam Data Validation Using a 1D Tank and Automated Couch Motions On a Ring-Gantry Linear Accelerator

N Knutson1,2*, M Schmidt1,2 , B Cai1 , E Laugeman1 , Y Hao1 , G Hugo1 , W Ngwa2 , S Mutic1 , H Li1 , E Sajo2 , (1) Washington University School of Medicine, St Louis, MO, (2) University of Massachusetts Lowell, Lowell, MA


(Tuesday, 7/16/2019) 3:45 PM - 4:15 PM

Room: Exhibit Hall | Forum 3

Purpose: Develop an independent beam data acquisition method for treatment planning system (TPS) validation using a 1D tank and automated couch motions on a ring-gantry linear accelerator.

Methods: Beam data acquisition using a 3D tank was first performed on a ring-gantry linear accelerator (Halcyon, Varian Medical Systems, Palo Alto CA), which came preconfigured. These measurements were then repeated using a TG-51 1D water tank with automated couch motions. These two different measurement geometries were then simulated using the TPS. Four data sets (1DS_measured, 1DS_simulated, 3DS_measured, 3DS_simulated) were then compared using 1D Gamma.

Results: All measured data yielded a high percentage of points with Γ(1%/1mm) < 1, 97.8% for 3DS and 98.8% for the 1DS, when compared to respective TPS simulated data. All data points yielded a Γ(2%/2mm) < 1 for the 3DS and Γ(1.5%/1.5mm) < 1 for the 1DS. Comparing across measurement geometries, 93.0% of points yielded Γ(1%/1mm) < 1 and 100% of points (Γ(3%/3mm) < 1), suggesting a slight difference between 1DS and 3DS, purely due to the difference in scanning geometry. This was confirmed by comparing simulations of the scanning geometries. The ratio of these simulations was used as a correction factor to cross-compare measurements between geometries. Correcting for scanning geometry, 1DS_corrected data matched 3DS_measured data with 99.0% of points yielding Γ(1%/1mm) < 1 and 100% of points Γ(1.5%/1.5mm) < 1.

Conclusion: The 1DS shows excellent agreement with the TPS and the 3DS, while providing significant logistical and cost savings. The small differences present between the two measurement geometries can be accurately predicted by the TPS. Suggesting, future TPS beam modelling could feasibly use the 1DS data if the TPS accounted for the measurement geometry used instead of assuming an infinite geometry by default. This method was demonstrated using Halcyon but could be extended to other treatment platforms.

Funding Support, Disclosures, and Conflict of Interest: Nels Knutson, Sasa Mutic, Geoff Hugo, Bin Cai, receive research funding from Varian Medical Systems. Sasa Mutic and Matt Schmidt receive consulting fees from Varian Medical Systems.


Dosimetry, Commissioning, Quality Assurance


TH- External beam- photons: Development (new technology and techniques)

Contact Email