319 / 1224
How efficient can MC be …?
t
CALC
scales with N, type A uncertainty scales with
➢
The only means to decrease t
CALC
for a given level of type A uncertainty is
variance reduction.
➢
Variance reduction:
•
Simple techniques: geom. truncation, E cut off, phase space files*, analytical
primary scatter separation (PSS), …
(* Pantelis et al. On source models for 192Ir HDR brachytherapy dosimetry using model based algorithms
Phys. Med. Biol. 61, 2016)
•
Elaborate techniques**: techniques to bias the sampling distributions while using
correction factors to eliminate the biasing effect in the sample mean of the quantity
of interest
(** see Sheikh-Bagheri, D., Kawrakow, I., Walters, B., and Rogers, D.W.O. 2006. Monte Carlo simulations:
Efficiency improvement techniques and statistical considerations. Integrated New Technologies into the Clinic:
Monte Carlo and Image-Guided Radiation Therapy—Proceedings of the 2006 AAPM Summer School)