Monte Carlo simulations for configuring and testing an analytical proton dose-calculation algorithm

• Published in: Physics in medicine & biology Vol. 52; no. 15; p. 4569
• Main Authors: Newhauser, Wayne, Fontenot, Jonas, Zheng, Yuanshui, Polf, Jerimy, Titt, Uwe, Koch, Nicholas, Zhang, Xiaodong, Mohan, Radhe
• Format: Journal Article
• Language: English
• Published: England 07-08-2007
• Subjects: Algorithms; Computer Simulation; Models, Biological; Monte Carlo Method; Neoplasms - radiotherapy; Protons - therapeutic use; Radiometry - methods; Radiotherapy Dosage; Radiotherapy Planning, Computer-Assisted - methods; Radiotherapy, Conformal - methods; Reproducibility of Results; Sensitivity and Specificity; Software Validation
• ISSN: 0031-9155
• DOI: 10.1088/0031-9155/52/15/014

Contemporary treatment planning systems for proton radiotherapy typically use analytical pencil-beam algorithms - which require a comprehensive set of configuration data - to predict the absorbed dose distributions in the patient. In order to reduce the time required to prepare a new proton treatment planning system for clinical use, it was desirable to configure the planning system before measured beam data were available. However, it was not known if the Monte Carlo simulation method was a practical alternative to measuring beam profiles. The purpose of this study was to develop a model of a passively scattered proton therapy unit, to simulate the properties of the proton fields using the Monte Carlo technique and to configure an analytical treatment planning system using the simulated beam data. Additional simulations and treatment plans were calculated in order to validate the pencil-beam predictions against the Monte Carlo simulations using realistic treatment beams. Comparison of dose distributions in a water phantom revealed small dose difference and distances to agreement under the validation conditions. The total simulation time for generating the 768 beam configuration profiles was approximately 6 weeks using 30 nodes in a parallel processing cluster. The results of this study show that it is possible to configure and test a proton treatment planning system prior to the availability of measured proton beam data. The model presented here provided a means to reduce by several months the time required to prepare an analytical treatment planning system for patient treatments.