Investigation of Dose Thresholds and Normalization Methods Effect on Gamma Index Analysis for SRT and SBRT Patients with a Monte Carlo Secondary Dose Check Software

Abstract

PurposeWe aimed to analyze the effects of different dose thresholds (DTHs) and normalization methods on 3-dimensional (3D) gamma index analysis (GIA) for SRT/SBRT using Monte Carlo (MC)-based secondary dose check quality assurance (SDC-QA) software.MethodsThirty VMAT plans (brain SRT and lung SBRT) were calculated using treatment planning system algorithms (TPS) and recalculated in MC-based SDC-QA software. Dose comparisons between TPS algorithms and MC were performed with 3D-GIA utilizing global and local normalization methods. Different DTHs were used to analyze their effects on the gamma passing rates (GPRs) of the plans (GPRs(Plan)). Target-specific GPRs (GPRs(PTV)) were calculated to investigate the relationship between GPRs(Plan) and GPRs(PTV). The results obtained by two normalization methods in low and high-dose regions were statistically compared.ResultsGPRs(Plan) increases in global normalization and decreases in local normalization when DTHs decrease. GPRs(Plan) calculated with DTH of 5% and 10% were > 95% even if GPR(PTV) was > 95% in global normalization. GPR(Plan) was < 95% with DTHs of 5% and 10% in 56 of 62 evaluations even when GPR(PTV) was > 95% in local normalization. Between global and local normalizations, there were significant differences (p < 0.05) in the low dose region whereas differences were insignificant (p > 0.05) in the high dose region.ConclusionWhen 3D-GIA is unavailable, DTHs of 5% and 10% should be used with caution in SRT/SBRT patients. They may mask errors in low-dose regions with global normalization and cause us to miss an acceptable match in high-dose regions with local normalization. Both normalization methods may be preferred for this group of patients in the high dose region.