Verification of output factors for small photon beams using Monte Carlo methods

In order to achieve acceptable radiosurgery delivery with small photon beams, it is important to know the radiation characteristics of the beams produced. One problem is that the finite volume of standard detectors (ion chambers, solid state detectors) causes uncertainties in measurements of paramet...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bezrukiy, N.V., DeMarco, J.J., Chetty, I., Smathers, J.B., Solberg, T.D.
Format: Tagungsbericht
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In order to achieve acceptable radiosurgery delivery with small photon beams, it is important to know the radiation characteristics of the beams produced. One problem is that the finite volume of standard detectors (ion chambers, solid state detectors) causes uncertainties in measurements of parameters for small beams. This investigation presents a computational verification of measurements of dosimetric parameters of small photon beams. Circular fields (as small as 3 mm in diameter) were simulated using MC methods to verify beam characteristics. The standard release of the MCNP4b (Monte Carlo N-Particle, version 4b) software from Los Alamos National Lab was used to perform the calculations. Due to very small tally volumes, large numbers of histories were needed to achieve acceptable statistics on our results. Parallel computing techniques, utilizing the Parallel Virtual Machine (PVM) software, were employed to reduce calculation times. Originally developed- for Unix platforms, the authors have implemented PVM on a cluster of non-dedicated, general purpose PCs running Windows NT. This serves as a low cost, high performance computing resource. Output factors calculated with MCNP agreed within 2.05% (maximum for any collimator size) with those determined by direct measurement for all field sizes above 7 mm in diameter. Below this value, larger discrepancies were observed and Monte Carlo is thought to be the superior technique in determining the output factors. Parallel processing using PVM proved to be a very useful tool in reducing the calculation time of the MCNP4b code.
ISSN:1094-687X
1558-4615
DOI:10.1109/IEMBS.2000.900441