Clinical Implementation of Log File Analysis for Patient-Specific QA
Abstract
Purpose
To clinically validate log file–based patient-specific quality assurance (PSQA) for IMRT and VMAT treatments as an efficient alternative to physical measurement-based methods, while maintaining compliance with AAPM recommendations.
Methods
Radformation Log File Analysis v2.5 (LFA), integrated within Eclipse TPS v16.12, was evaluated using the RadMonteCarlo dose reconstruction engine. A representative cohort of twenty-two IMRT and VMAT plans was delivered on two Varian TrueBeam linacs. Trajectory log files capturing delivered MLC positions, gantry motion, and dose rate were analyzed. Consistent with TG-218 guidance, gamma analysis was performed using clinically relevant criteria (distance to agreement of 2mm and 3% dose difference at 3mm voxel resolution) to evaluate both whole-dose matrix and high-dose PTV performance. Alternate evaluation criteria included voxel size reduction to 1.5mm and/or relaxed dose-difference threshold up to 5% for further assessment.
Results
LFA demonstrated reasonable agreement between planned and delivered dose. Eleven cases achieved >95% gamma passing for the full dose matrix and >90% for high-dose PTVs using standard criteria. Cases involving air–tissue interfaces, heterogeneous anatomy, buildup regions, or implanted hardware demonstrated lower initial pass rates, consistent with known algorithmic differences between AAA and Monte Carlo dose calculations. The lowest result, 93.03% for the high dose PTV, was for the target of a head and neck treatment site, which was near an airway and near the surface at the buildup region, requiring the alternate gamma criteria. Comparable trends were observed between LFA and conventional measurement-based QA for complex cases.
Conclusion
In alignment with AAPM TG-218 and supported by routine TG-142 machine QA, LFA provides a robust, efficient, and clinically viable approach for PSQA. When supported by clearly defined acceptance criteria and continued machine QA, LFA enables streamlined workflows while preserving dosimetric confidence, particularly when used with informed clinical judgment for heterogeneous treatment sites.