基于三维Savitzky-Golay滤波的蒙特卡罗剂量分布去噪

De-Noising of Monte Carlo Dose Distribution Based on Three-Dimensional Savitzky-Golay Filter

提出了一种基于三维Sav itzky-G o lay滤波的蒙特卡罗(M C)剂量分布的去噪方法。该方法首先利用M C方法模拟粒子轨迹数目较少时得到剂量的三维分布,然后对该剂量分布用三维Sav itzky-G o lay平滑滤波方法进行去噪处理。结果表明:采用三维Sav itzky-G o lay平滑滤波方法去噪,不仅提高了剂量分布的可视性,降低了M C计算剂量分布的不确定性,而且也相应地提高了M C剂量计算方法的计算效率。

A de-noising algorithm for Monte Carlo （MC） simulated dose distributions is presented. The algorithm is based on three-dimensional Savitzky-Golay filter. Firstly, three-dimensional dose distributions are simulated by MC method with fewer particles, then the Savitzky-Golay filter is applied to the dose distributions for de-noising. The results indicate that using Savitzky-Golay smoothing filter to de-noising can improve the visualizability of the dose distribution, reduce the uncertainty of MC simulated dose distributions, and increase the calculated efficiency of MC dose distributions.