Increasing efficiency of BEAMnrc‐simulated Co‐60 beams using directional source biasing. Issue 10 (15th September 2015)

• Record Type: Journal Article
• Authors: Walters, B. R. B.
• Is Part Of: Medical physics. Volume 42:Issue 10(2015)
• Page Start: 5817
• Page End: 5827
• Publication Date: 2015-09-15
• Subjects: C++ language -- cobalt -- dosimetry -- Monte Carlo methods -- phantoms -- radiation therapy; Applications -- Applications of Monte Carlo methods -- Monte Carlo simulations -- Dosimetry/exposure assessment -- Therapeutic applications, including brachytherapy -- Dose‐volume analysis; Radiation therapy -- Scintigraphy; Co‐60 beams -- Monte Carlo simulation -- BEAMnrc -- variance reduction techniques; Photons -- Schottky barriers -- Contaminants -- Electron beams -- Collimators -- Ion beam sources -- Dosimetry -- Monte Carlo methods -- Secondary emission -- Ionization chambers; Medical physics -- Periodicals; Medical physics; Geneeskunde; Natuurkunde; Toepassingen; Biophysics; Periodicals; Periodicals; Electronic journals; 610.153
• DOI: 10.1118/1.4930060
• Languages: English
• ISSNs: 0094-2405

Abstract

Abstract : Purpose: This study describes the implementation of a directional source biasing (DSB) scheme for efficiently simulating Cobalt‐60 treatment heads using the BEAMnrc Monte Carlo code. Previous simulation of Co‐60 beams with BEAMnrc was impractical because of the time required to track photons not directed into the treatment field and to simulate secondary charged particles. Methods: In DSB, efficiency is increased by splitting each photon emitted by the Co‐60 source a user‐defined number of times. Only those split primary photons directed into a user‐defined splitting field (encompassing the treatment field) are sampled, yielding many low‐weight photons directed into the field. Efficiency can be further increased by taking advantage of radial symmetry at the top of the treatment head to reduce the number of split primary photons tracked in this portion. There is also an option to generate contaminant electrons in DSB. Results: The DSB scheme in BEAMnrc increases the photon fluence calculation efficiency in a 10 × 10 cm 2 Co‐60 beam by a factor of 1800 with a concurrent increase in contaminant electron fluence calculation efficiency by a factor of 1200. Implementation of DSB in beampp, aC++ code for accelerator simulations based on EGSnrc and theC++ class library, egspp, increases photon fluence efficiency by a factor of 2800 and contaminant electron fluence efficiency by a factor of 1600. Optimum splitting numbers are in the range of 20 000–40 000. For dose calculations in a water phantom (0.5 × 0.5 × 0.5 cm 3 voxels) this translates into a factor of ∼400 increase in dose calculation efficiency (all doses > 0.5 × D max ). An example calculation of the ratio of dose to water to dose to chamber (the basis of the beam quality correction factor) to within 0.2% in a realistic chamber using a full simulation of a Co‐60 treatment head as a source indicates the practicality of Co‐60 simulations with DSB. Conclusions: The efficiency improvement resulting from DSB makes Monte Carlo commissioning of Co‐60 beams and calculation of beam quality correction factors feasible.