Premium
TH‐F‐211‐03: A More Accurate Dose Calculation for Volumetric Modulated Arc Therapy
Author(s) -
Yang J,
Zhang P,
Hunt M,
Lim S,
LoSasso T,
Mageras G
Publication year - 2012
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 - 10.1118/1.4736399
Subject(s) - resolution (logic) , aliasing , imaging phantom , aperture (computer memory) , mathematics , transmission (telecommunications) , sampling (signal processing) , algorithm , nuclear medicine , physics , filter (signal processing) , optics , computer science , acoustics , detector , medicine , artificial intelligence , computer vision , telecommunications
Purpose: Dose discrepancy emerges between the calculation algorithm and measurements due to MLC tongue and groove (TG) effect, and rapid changes of the MLC apertures between adjacent control points (CP) in hypofractionated Volumetric Modulated Arc Therapy (VMAT) that demands high dose modulation. This study explores methods to correct these dose calculation errors. Methods: Because the physical width of tongue or groove is around 1mm, the 2mm transmission matrix resolution used in the current calculation algorithm may result in errors due to aliasing in the MLC transmission sampling. A new calculation algorithm using 1mm resolution is implemented to remove aliasing and more accurately model TG transmission. The commonly used gantry angle sampling resolution of 2 degree causes significant error when there is large MLC movement from one CP to next. Instead of inserting additional CP which doubles the calculation time, we insert two more apertures that are interpolated at a 0.7 degree resolution before and after each CP. The weighted average transmission is then used for calculation. Dose calculations are compared with film measurements for hypofractionated VMAT in lung, hip, prostate and paraspinal cases to assess accuracy. Results: Up to 10% maximal local and 3.3% average field dose discrepancy exist with the 2mm resolution algorithm. The errors are reduced to 5% maximal and 2% average with the 1mm resolution algorithm, and further reduced to within 3% maximal and 0.5% average using the finer aperture angular resolution. With both corrections built in, the gamma (3%,3mm) pass rate is closer to 100% with improvement ranging from 2% to 14% relative to pass rate without correction. Conclusions: Finer transmission and angular resolution needs to be implemented in the dose calculation algorithm to handle the highly modulated fields in hypofractionated VMAT and consistently achieve a high accuracy required by the QA program.