Purpose: Analysis of trajectory logs (TL) is increasingly used as an efficient method of patient specific IMRT QA. Mechanical and dosimetric linac parameters recorded during plan delivery can be extracted from TL and used by independent software to calculate the delivered 3D dose distribution on the planning CT to compare to the planned dose. The purpose of this study was to investigate the use of TL for IMRT QA of Lung SBRT plans.
Methods: Twenty clinical lung SBRT plans were retrospectively analyzed. Conventional IMRT QA was performed by delivering the plans to the Sun Nuclear SRS MapCheck, and a Gamma analysis (3%,1mm) was performed. TL of the QA delivery were analyzed in Sun Nuclear’s PerFraction, where the 3D delivered dose was calculated using the SDC collapsed cone algorithm and compared to the Eclipse planned dose calculated with both AAA and AcurosXB algorithms. 3D Gamma analysis comparisons used 3%,1mm, 40% threshold. MLC RMS errors were calculated from the TL. Pearson correlation was calculated for PerFraction and MapCheck Gamma results.
Results: MapCheck average (±SD) Gamma pass rate (GPR) was 98.3±1.3% for AAA and 99.1±1.3% for AcurosXB. The average MLC RMS and 95th percentile errors were 0.022mm and 0.033mm, respectively, indicating no appreciable leaf errors during the delivery. TL PerFraction average 3D GPR for AAA was 91.1±10.1%, with seven plans below 90%. Results improved for AcurosXB, with an average GPR of 99.3±1.3%, and no plans below 95%. MapCheck and PerFraction results were not significantly correlated.
Conclusion: TL QA results were dependent on the TPS dose algorithm used in the original plan. Results support previous literature on the improved accuracy of collapsed cone and AcurosXB vs AAA for lung treatments. When using AcurosXB, TL can be safely used for patient-specific QA of lung SBRT plans based on agreement with SRS MapCheck measurements.