Premium
Monte Carlo evaluation of the convolution/superposition algorithm of Hi‐Art™ tomotherapy in heterogeneous phantoms and clinical cases
Author(s) -
Sterpin E.,
Salvat F.,
Olivera G.,
Vynckier S.
Publication year - 2009
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.3112364
Subject(s) - imaging phantom , monte carlo method , tomotherapy , convolution (computer science) , dosimetry , superposition principle , nuclear medicine , histogram , materials science , physics , computational physics , mathematics , computer science , optics , radiation therapy , statistics , medicine , mathematical analysis , radiology , machine learning , artificial intelligence , artificial neural network , image (mathematics)
The reliability of the convolution/superposition (C/S) algorithm of the Hi‐Art™ tomotherapy system is evaluated by using the Monte Carlo model TomoPen, which has been already validated for homogeneous phantoms. The study was performed in three stages. First, measurements with EBT Gafchromic film for a 1.25 × 2.5cm 2field in a heterogeneous phantom consisting of two slabs of polystyrene separated with Styrofoam were compared to simulation results from TomoPen. The excellent agreement found in this comparison justifies the use of TomoPen as the reference for the remaining parts of this work. Second, to allow analysis and interpretation of the results in clinical cases, dose distributions calculated with TomoPen and C/S were compared for a similar phantom geometry, with multiple slabs of various densities. Even in conditions of lack of lateral electronic equilibrium, overall good agreement was obtained between C/S and TomoPen results, with deviations within 3%/2 mm, showing that the C/S algorithm accounts for modifications in secondary electron transport due to the presence of a low density medium. Finally, calculations were performed with TomoPen and C/S of dose distributions in various clinical cases, from large bilateral head and neck tumors to small lung tumors with diameter of < 3 cm . To ensure a “fair” comparison, identical dose calculation grid and dose‐volume histogram calculator were used. Very good agreement was obtained for most of the cases, with no significant differences between the DVHs obtained from both calculations. However, deviations of up to 4% for the dose received by 95% of the target volume were found for the small lung tumors. Therefore, the approximations in the C/S algorithm slightly influence the accuracy in small lung tumors even though the C/S algorithm of the tomotherapy system shows very good overall behavior.