A simplified methodology to produce Monte Carlo dose distributions in proton therapy

Abstract

The purpose of this study was to develop a simplified methodology that will produce Monte Carlo (MC) dose distribution for proton therapy which can be used as a clinical aid in determining the adequacy of proton plans produced from the treatment planning system (TPS). The Geant4 Monte Carlo toolkit was used for all simulations. The geometry of the double scatter nozzle in the simulation was a simplification of the treatment nozzle. The proton source was modeled as discrete energy layers, each with a unique energy distribution and weighting factor. The simplified MC system was designed to give the same dose distribution as the measured data used to commission the TPS. After the simplified MC system was finalized, a series of verification comparisons were made between it, measurements, and the clinically used TPS. Comparisons included the lateral profile of a stair‐shaped compensator that simulated a sharp lateral heterogeneity and depth‐dose measurements through heterogeneous materials. The simplified MC system matched measurements to within 2% or 2 mm for all commissioning data under investigation; moreover, the distal edge and lateral penumbra was within 1 mm of the measurements. The simplified MC system was able to better reproduce the measured profiles for a stair‐shaped compensator than the TPS. Both MC and TPS matched the measured depth dose through heterogeneous materials to within 2% or 2 mm. The simplified MC system was straightforward to implement, and produced accurate results when compared to measurements. Therefore, it holds promise as a clinically useful methodology to validate the relative dose distribution of patient treatment plans produced by the treatment planning systems.PACS number: 87.55.K‐, 87.55.ne

Clinical Validation of a Ray-Casting Analytical Dose Engine for Spot Scanning Proton Delivery Systems

“…A fast Graphic-Processing Unit (GPU)-accelerated MC dose calculation engine developed at Mayo Clinic, Rochester, Minnesota 13,14 and a commercial TPS, Eclipse version 13.7 (Varian Medical Systems, Palo Alto, California), 38 were used to assess the relative accuracy of the DE1. Henceforth, for simplicity, the abbreviation DE1 refers to the analytical dose engine, MC2 refers to the fast MC code, and TPS3 refers to the commercial TPS.…”

“…Monte Carlo simulation calculates the field size effect and inhomogeneity problems directly from the underlying physics, but even fast MC methods 13,14 are currently not widely available or fast enough for iterative optimization. Although MC dose calculations are suggested when treating highly heterogeneous sites, 10 –12 comprehensively commissioned and validated analytic dose calculation models 6,7 provide higher computational speed required for many routine clinical tasks.…”

Highly efficient and sensitive patient-specific quality assurance for spot-scanned proton therapy

“…The profiles are a function of the beam energy and the distance along the beam central axis [38]. While proton fluence in spot-scanning systems can be well-modeled, the challenge of proton patient specific QA arises in accurately calculating the dose to the patient in the presence of heterogeneities, which affect the proton range and lateral scattering conditions [16,17,18]. Although analytic algorithms perform acceptably in many scenarios, the plan-specific nature of proton beams incident on heterogeneity interfaces means that a second check of the primary, analytic dose calculation algorithm is absolutely a necessary patient-specific quality check.…”

“…Proton plans calculated with pencil beam convolution algorithms, however, are susceptible to particle range uncertainties and to distortions in dose distributions, primarily due to the incorrect modeling of multiple scattering in inhomogeneous media. On the other hand, the Monte Carlo method handles particle transport in heterogeneous patient geometries more accurately, and is widely accepted to be the “gold standard” technique for dose calculation [16,17,18]. The limitations of analytic proton dose calculations have previously been reported by several authors.…”

A Data-Driven Fragmentation Model for Carbon Therapy GPU-Accelerated Monte-Carlo Dose Recalculation

“…The advent of general programming Graphics Processing Units (GPU) has prompted the development of MC algorithms that can significantly reduce the plan recalculation time (10)(11)(12)(13)(14)(15)(16)(17)(18)(19) achieving an impressive speed gain compared to CPU-based calculations, profiting from algorithmic simplifications and hardware acceleration. Exploiting the GPU hardware, many vended TPS used for proton therapy now include MC tools (20)(21)(22)(23)(24)(25). For carbon therapy, recently a tool called goCMC (GPU OpenCL Carbon Monte Carlo) (26) was developed.…”