Post-Treatment Quality Assurance of Real-Time Adaptation Deliveries
Abstract
Purpose
To develop a method for verifying real-time adaptation treatment log data generated by the Accuray Radixact Synchrony system, for post-treatment quality assurance as recommended by the AAPM TG264 report.
Methods
An in-house software system was developed to parse and analyze patient treatment planning and delivery data from Accuray Patient Data Extractor archives. Treatment gantry angle and IEC-X and IEC-Z target positions were correlated against delivery time to calculate lateral displacement of target in beams-eye view, relative to planned position. Predicted leaf adaptations, where the displacement exceeded adaptation thresholds, were then evaluated against differences between planned and recorded leaf sinograms and differences between planned and exit detector measured leaf sinograms. The in-house software produces leaf sinogram figures and summary statistics describing expected and logged fraction of treatment with adaptation, and indicates any discrepancies with leaf open times across gantry angle projections.
Results
For many treatment fractions, the evaluation of beam’s-eye-views of target motion alongside visualisations of MLC adaptation allowed occasions where adaptation occurred to be correlated with periods when targets moved beyond the adaptation threshold, although some cases are still under investigation. For four of the 1,366 Synchrony treatment fractions so far evaluated, anomalous multi-leaf collimator (MLC) motion was identified in the delivered sinograms, indicating occasions where individual Radixact MLC leaves had remained open (for less than one gantry rotation) and continued adapting when no longer directed at the treatment target, leading to minor (<0.5% of prescription) deviations from the planned dose over small regions of tissue.
Conclusion
This retrospective investigation of the real-time motion adaptation achieved by 1,366 Radixact Synchrony treatment fractions exemplifies the importance of post-treatment quality assurance of adaptive radiotherapy by highlighting the rare cases where issues with the handling of MLC adaptation were observed.