Abstract:
The purpose of this study was to evaluate the factors influencing electron dose calculation in Eclipse electron Monte Carlo (eMC). The dosimetric accuracy in the planar dose was assessed by MapCHECK2 and the point dose was assessed by the plane-parallel ionization chamber type PPC40 in solid water phantom. The factors influencing electron dose calculation in Eclipse eMC were evaluated by virtual phantoms (homogeneity, heterogeneity) in various conditions (irregular surface and extended SSD). Each parameter of accuracy, calculation grid size, random number generator seed, number of particle histories, smoothing method and smoothing level in Eclipse eMC was varied while other parameters were fixed in default value. Gamma analysis software was used to compare the dose distribution between the best parameter plan and vary parameter plans with 2%/2mm criteria. The 6 and 18 MeV energies with the sizes of 6×6 cm2 and 10×10 cm2 applicators were used. The measurement results of 18 MeV agreed with eMC calculation with higher than 90% gamma passing rate. The point dose measurement also showed the good agreement results with eMC. The accuracy, calculation grid size, smoothing method, smoothing level and number of particle history parameters had significant impact on the dose calculation. The average gamma passing rate of all parameter variations in homogeneity phantom with standard 100 cm SSD, homogeneity phantom with extended 110 cm SSD, heterogeneity phantom with material like bone, heterogeneity phantom with material like air, protrusion irregular shape surface phantom and depression irregular shape surface phantom were 74.2±17.7%, 82.3±21.1%, 80.7±21.8%, 66.3±23.3%, 56.3±19.5%, and 71.1±19.0%, respectively. The various situations had obvious impact on dose distribution. The heterogeneity, like air, and the irregular shape surface were presented in low gamma passing rate. The 6×6 cm2 applicator size showed lesser impact in dose distribution than larger size. The lower energy of 6 MeV presented the lesser impact in dose distribution than higher energy of 18 MeV. For the accuracy parameter, the 2% accuracy can be used instead of 1% that can reduce calculation time 3 times compared with 1% accuracy. For the calculation grid size, 2.5 mm is the optimal option in almost all situations with good agreement of gamma passing rate and 20 times reduction in calculation time compared with 1.0 mm grid size, but fine calculation grid size of 1.0 mm or 1.5 mm should be used in irregular shape surface. The 2D Median smoothing method can be applied in all situations except low energy with irregular shape that should be smoothed with 3D Gaussian method. For the smoothing level, the medium and strong smoothing level is recommended to select, except the case with air inhomogeneity in tissue and irregular body shape that should be calculated on strong smoothing level. The random generator seed number changing is not affected by dose distribution and calculation time. The number of particle histories other than 0 is not recommended.