Software commissioning of a Monte-Carlo model for the double-scattering treatment head at University Proton Therapy Dresden

Compared to treatment planning systems (TPS) in proton therapy, Monte-Carlo simulations have the potential to describe radiation fields in patients more precisely. However, next to an accurate Monte-Carlo model of the treatment head a benchmarking with respect to dose measurements is required. The purpose of this work was to set up and validate a Monte-Carlo simulation model of the clinical proton treatment fields at the University Proton Therapy Dresden (UPTD).
A detailed model of the treatment head geometry of the UPTD in double-scattering mode was implemented using the Monte-Carlo simulation environment TOPAS. The proton beam source was optimized to match measured reference depth-dose distributions for all clinically available treatment field options. The commissioned software model was validated against an independent set of depth-dose and lateral dose validation data measured in a water phantom. A setup for the direct simulation of so-called monitor units (MU), relating dosimeter readings in the treatment head to absolute dose in a water phantom, was implemented.
Validation data on depth-dose distributions were reproduced within range differences of 0.26 mm and a relative dose uncertainty of 1% for all treatment options (i.e., comparable to measurement uncertainties). Simulated lateral dose profiles differed from validation data in lateral width and penumbra less than 0.95 mm and 0.56 mm, respectively. Measured MU values were predicted within 2% accuracy for several reference and patient treatment fields.
The commissioned Monte-Carlo model reproduced the dose validation data measured during the clinical validation of the UPTD within clinical tolerances. This enables high-precision simulations of clinical proton beams. Furthermore, it has the potential to predict absolute doses and to use patient-specific MU values, which might decrease the patient-specific measurement effort at UPTD.
This, however, requires further validation.