Abstract
BackgroundMany dose calculation algorithms for radiotherapy planning need to be configured for each clinical beam using pre-defined measurements. An optimization process adjusts the physical parameters able to estimate the energy released in the medium in any geometrical condition. This work investigates the impact of measured input data quality on the configuration of the type “c” Acuros-XB dose calculation algorithm in the Eclipse (Varian Medical Systems) treatment planning system.MethodsDifferent datasets were acquired with the BeamScan water phantom (PTW) to configure 6 MV beams, for both flattened (6X) and flattening filter free mode (6FFF) for a Varian TrueBeam: (i) a correct dataset measured using a Semiflex-3D ion chamber, (ii) a set in missing lateral scatter conditions (MLS), (iii) a set with incorrect effective point of measurement (EPoM), (iv) sets acquired with PinPoint-3D chamber, DiodeP, microDiamond detectors.The Acuros-XB dose calculation algorithm (version 15.6) was configured using the reference dataset, the sets measured with the different detectors, with intentional errors, and using the representative beam data (RBD) made available by the vendor. The physical parameters obtained from each optimization process (spectrum, mean radial energy, electron contamination), were analyzed and compared. Calculated data were finally compared against the input and reference measurements.ResultsConcerning the physical parameters, the configurations presenting the largest differences were the MLS conditions (mean radial energy) and the incorrect EPoM (electron contamination). The calculation doses relative to the input data present low accuracy, with mean differences > 2% in some conditions. The PinPoint-3D ion chamber presented lower accuracy for the 6FFF beam. Regarding the RBD, calculations compared well with the input data used for the configuration, but not with the reference data.ConclusionThe MLS conditions and the incorrect setting of the EPoM lead to erroneous configurations and should be avoided. The choice of an appropriate detector is important. Whenever the representative beam data is used, a careful check under more clinical geometrical conditions is advised.
Highlights
In the radiotherapy process, the accuracy of the dose distribution calculation is fundamental
This rise brings the DiodeP and the missing lateral scatter conditions (MLS) to have the highest and lowest energy, respectively, after the filter, consistent with the measurements: the largest field Percentage depth dose (PDD) acquired in MLS conditions is less penetrating, while diode presents a higher penetration behavior (Supplementary Material)
For the 6FFF, the mean energy increase is less than 5% for the majority of the configurations, while the MLS presents a mean energy reduction of 10–20%
Summary
The accuracy of the dose distribution calculation is fundamental. In the 1980s, Brahme [1, 2] evaluated a variation of 6–7% in NTCP for a 1% change in delivered dose. It has already been pointed out in the ICRU Report 24 [3] that in critical situations, a dose calculation accuracy of ±5% must be achieved, but ±2% may be required. In 1993, Van Dyk et al [4] suggested acceptability criteria for photon beams, with ± 2% as an achievable goal for dose calculation algorithms accuracy. Many dose calculation algorithms for radiotherapy planning need to be configured for each clinical beam using pre-defined measurements. This work investigates the impact of measured input data quality on the configuration of the type “c” Acuros-XB dose calculation algorithm in the Eclipse (Varian Medical Systems) treatment planning system
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
