Get the Most Out Of Your Pre-treatment QA With Monte Carlo Accuracy
Monte Carlo is a gold standard throughout the industry for dose calculation, with reported accuracy that is supported by numerous publications, conferences, and workshops dedicated to the topic. In particular, the European Workshop on Monte Carlo Treatment Planning was held for several years (2006, 2009, and 2012) to stimulate scientific information exchange. I had the opportunity to attend two of the three workshops and still value the contacts and exchange of information provided. Since that time, the use of Monte Carlo in the clinic has steadily increased along with improved calculation speed.
In the US, the Imaging and Radiation Oncology Core (IROC) has been reporting for many years on dose inaccuracy when comparing the IROC phantom independent measurements to the clinic predictions. In a more recent publication, “IROC observed the surprising result that an independent recalculation dramatically outperformed institutional measurement-based QA in predicting whether a plan would pass the IROC phantom.” (Kry 2019)
IMSure 3D™ Software from Standard Imaging utilizes the SciMoCa Monte Carlo algorithm to perform an independent dose calculation. Using a proven dose calculation engine can provide improved error detection when compared to detector array measurements. Certainly, measurements do play an important role in QA, but perhaps more emphasis on machine-specific measurement QA and selective use of pre-treatment QA measurements is reasonable when evaluating an overall QA strategy. Increased use of in vivo rather than pre-treatment measurement-based methods for patient specific QA merits consideration. In addition, calculation-only pre-treatment QA is less labor intensive and does not require LINAC time. From automatic importing and calculating to generating reports ready for review within minutes, physicists can save time during their QA routine with IMSure 3D. Comprehensive analysis tools, which focus on relevant criteria such as target volume and OAR metrics to help efficiently assess differences, contribute to good QA.
Figure 2: TPS Dose, Monte Carlo Dose and Gamma Viewer
One of the sources of reported error stems from poor quality beam models, with the most recent publication stating “typical TPS beam modeling parameters are associated with failing phantom audits. This is identified as an important factor contributing to the observed failing phantom results, and highlights the need for accurate beam modeling.” (Glenn 2022)
Dose calculation engines of all algorithm types require high quality input in the form of the beam model. Beam modelling often uses the water tank scan data along with output factor measurements as a function of field size to tune each field independently. This data may contain inconsistencies that become 1:1 dose discrepancies of the same magnitude in the dose calculation. Custom vendor-provided beam modelling allows for comparison to data from other LINACs using the same energy.
A Monte Carlo beam model is consistent throughout, as the virtual sources rely on parameters (Ex: a primary spectra, secondary spectra and source sizes) that must be fit to perform well at all field sizes. The beam model accuracy is ensured in IMSure 3D through custom machine-specific beam models for highly accurate simulations of planned delivery.
Kry SF, Glenn MC, Peterson CB, et al. Independent recalculation outperforms traditional measurement-based IMRT QA methods in detecting unacceptable plans. Med Phys.2019;46(8):3700-3708.
Glenn MC, Brooks F, Peterson CB, Howell RM, Followill DS, Pollard-Larkin JM, Kry SF. Photon beam modeling variations predict errors in IMRT dosimetry audits. Radiother Oncol. 2022 Jan;166:8-14.