Article
Verification of Simplified Monte Carlo Algorithm in Treatment Planning for Proton Cancer Therapy
Search Medline for
Authors
Published: | September 24, 2009 |
---|
Outline
Text
Background: The accurate dose calculation in treatment planning is important for proton cancer therapy. Many facilities use the pencil beam dose calculation algorithm (PBA) or its variants. However, since it does not take into account the beam-path confusion in the heterogeneous human tissue, the accuracy is limited. Although full Monte Carlo calculations can produce more accurate dose distributions, it requires a long calculation time.
Kohno et al. have developed a Simplified Monte Carlo algorithm (SMC) to realize a fast and accurate dose calculation method [1]. We verify the effectiveness of the algorithm by comparing the calculation results with measured dose distributions in a combination of slab phantoms.
Material and methods: Calculation: The SMC traces the trajectories taking into account the multiple Coulomb scattering of individual protons using the Highland's formula. The range losses in the range compensator (RC) and the aperture are calculated using the water-equivalent thicknesses of individual materials. The relative dose deposit in a phantom voxel is obtained from the residual range in water of a proton using a measured Bragg curve in water.
Experiment: We verify the calculation accuracy by comparing it with measured proton dose distribution in heterogeneous phantoms using the modulated beam with a SOBP width of 80 mm of 150 MeV protons from the cyclotron in National Cancer Center Hospital East. We used the PTW 2D-Array seven29, 27 x 27 arrayed chambers with a sensitive volume of 5 mm x 5 mm x 5 mm. We designed the experimental arrangement for all protons to stop at z=60 mm as Figure 1 [Fig. 1]. The phantoms are made of Tough Water(TW), Tough Lung(TL), and Tough Bone(TB) phantoms, the water equivalent thickness ratio of which were 1.0, 0.33, 1.5, respectively.
Results: Figure 2 [Fig. 2] shows the comparison of measured and calculated lateral dose distributions at z=0, 60 mm. We notice that two high-dose region around x=+-50 mm at z=0 due to the scattered protons in the aperture edge. We also noticed a low-dose region around x =0 at the depth z=60 mm which is produced by protons incident on the thinner part of the RC, passing through the Tough bone region, and stopped before z=60 mm. While the SMC reproduces the two cases well, the PBA fails to reproduce them. The calculation time was 45 minutes for SMC(four million particles), 3 minutes for PBA, by the iMac with a CPU of 2.7 GHz Intel Core 2 Duo.
Conclusion: We observed that the SMC method reproduced well the measured dose distributions with a reasonable calculation time. The result comes from the fact that the algorithm takes into account the collimator edge scattering effect and the beam-path confusion in the heterogeneous phantoms.
References
- 1.
- Kohno R, Sakae T, Takada Y, et al. Simplified Monte Carlo Dose Calculation for Therapeutic Proton Beams. Jpn J Appl Phys. 2002;41:L294-L297. DOI: 10.1143/JJAP.41.L294