Detecting anomalies in a deliberately biased tomotherapy plan: Comparison of two patient-specific quality assurance processes involving ArcCHECK ? and Gafchromic ? EBT3 films

摘要

Abstract Purpose This work proposes a comparative evaluation of two of our patient-specific quality assurance processes involving ArcCHECK ? (Sun Nuclear) and Gafchromic ? EBT3 films (Ashland) in order to determine which detector is able to most effectively detect an anomaly in a deliberately biased tomotherapy plan. Material and methods A complex clinical head and neck tomotherapy plan was deliberately biased by introducing six errors: multileaf collimator leaf positional errors by leaving one and two central leafs closed during the whole treatment, initial radiation angle errors (+0.5° and +1.0°) and multileaf collimator leafs opening time errors (+0.5% and +1.0%). For each error-induced plan, comparison of ArcCHECK ? with Gafchromic ? EBT3 films (20.3×25.4cm 2 ) was performed through two methods: a dose matrices subtraction study and a gamma index analysis. Results The dose matrices subtraction study shows that our ArcCHECK ? processing is able to detect all the six induced errors contrary to the one using films, which are only able to detect the two biases involving multileaf collimator leaf positional errors. The gamma index analysis confirms the previous method, since it shows all six errors induced in the reference plan seem to be widely detected with ArcCHECK ? with the more restrictive 1%/1mm gamma criterion, whereas films may only be able to detect biases in relation to multileaf collimator leaf positional errors. It also shows the common 3%/3mm gamma criterion does not allow deciding between both detectors in the detection of the six induced biases. Conclusion Both comparative methods showed ArcCHECK ? processing is more suitable to detect the six errors introduced in the reference treatment plan.