Monte Carlo simulation techniques are widely used in medical physics. One can distinguish between general purpose Monte Carlo codes which can be applied to a wide range of problems and Monte Carle codes that are specialised for specific purposes. VERIQA applies the SciMoCa^TM Monte Carlo code which is a specialised code developed for the efficient calculation of radiation treatment dose.

For dose calculation, the Monte Carlo principle is used to predict the absorbed energy by simulating the individual transport of particles passing through matter. Monte Carlo therefore applies the knowledge of the patient geometry derived from imaging and the physical knowledge of particle interactions gained from experimental data. The sophisticated simulation of physical interactions is the reason why Monte Carlo is often considered the gold standard for dose calculation in radiation therapy.

A crucial aspect of accurate Monte Carlo simulations is the precise definition of the particle origin. In radiotherapy, the particle origin is described in linac-specific beam models that simulate the accelerator head.  

VERIQA RT MonteCarlo 3D applies is a specialized Monte Carlo code derived from the EGSnrc/XVMC/VMC++ Monte Carlo code family. It keeps deviations in the toughest artificial situations to a maximum of 2% compared to general purpose Monte Carlo codes. Due to its unique virtual beam modelling and optimized use of sophisticated variance reduction techniques, it maximizes efficiency and minimizes non-Gaussian noise.

VERIQA does not use GPU base dose calculation.  For the efficient simulation of the accelerator head and variable collimators, variance reduction techniques can be applied to great benefit. Variance reduction techniques come with an overhead in terms of memory management, and with some price in terms of code complexity. Both factors are a greater problem for a GPU architecture, which is specialized for executing the same set of instructions in parallel on many data entities (SIMD, single instruction, multiple data) than for a CPU architecture (MIMD, multiple instructions, multiple data). VERIQA is designed to be versatile and work with very diverse accelerator head geometries. It has a finely balanced suite of variance reduction techniques to couple the accelerator head model with the patient model. Therefore, the code executes preferentially on a CPU architecture.

VERIQA`s Monte Carlo algorithm was extensively benchmarked against EGSnrc. It was validated for materials between mass densities
of 0.25 g/cm³ and 8.0 g/cm³ and photon energies between 1 MeV and 25 MeV. The validation of the linac beam models was based on dose measurements under controlled conditions. Furthermore, comparison to measurements of realistic treatment plans served the purpose of comprehensive end-to-end testing.

Each VERIQA linac head is simulated with a type- and energy- specific beam model. VERIQA beam models are created by PTW experts based on water phantom and MLC calibration measurements. Beam models can either be provided as standardized or customized beam models.