Efficient Monte Carlo simulations in kilovoltage x-ray beams

Kilovoltage x-ray systems are modeled with BEAMnrc using directional bremsstrahlung splitting, which is five to six orders of magnitude more efficient than a simulation without splitting and 60 times more efficient than uniform bremsstrahlung splitting. Optimum splitting numbers are between 2 and 3 orders of magnitude larger than for megavoltage beams.;In the medium-energy range, calculations reproduce experimental half-value layer values to better than 2.3%. For mammography beams a difference of 0.5% and 2.5% with experiment is obtained with and without a scaling of the tungsten L-shell EII cross-sections respectively. For low-energy lightly-filtered beams a scaling factor of 2.1 gives the best agreement (∼ 3%) with the experiment, significantly worsening to 8% for a scaling factor of 1.8, which gives the best match for Aatt.;The fast algorithm for calculating the scatter contribution to cone beam computed tomography scans increases the efficiency by more than 3 orders of magnitude. Smoothing the scatter distribution pushes the efficiency gain over four orders of magnitude. The iterative correction algorithm removes the scatter from the measured scans improving the accuracy of the reconstructed image. The dependence of image reconstruction accuracy on the sophistication of the photon interaction models is investigated. No significant difference is observed when using models including coherent and incoherent scattering. Considering only incoherent scattering for free electrons shows a significant bias.;A self-consistent approach for the calculation of free-air chamber correction factors with the EGSnrc Monte Carlo system is introduced. In addition to the traditional factors employed to correct for attenuation (A att), photon scatter (Ascat) and electron energy loss (Aeloss), correction factors for aperture leakage (Aap) and backscatter (A b) are defined. Excellent agreement is obtained between calculated and measured Ascat and Aeloss values. Computed Aatt values for medium-energy and mammography beams reproduce the measurements well. For low-energy lightly-filtered beams, Aatt values show significant differences with the experiment. Scaling the tungsten L-shell EII cross-sections by a factor of 2 eliminate these differences. The inconsistency of the evacuated-tube technique for measuring Aatt is negligible for medium-energy and mammography beams, and 0.2% for low-energy lightly-filtered beams The aperture correction Aap becomes significant in the medium-energy range with increasing energy. The newly introduced backscatter correction Ab becomes as high as 0.4% in the low-energy range.