Room: AAPM ePoster Library
Purpose: To precisely evaluate dose discrepancy associated with beam delivery and treatment planning parameters in patient-specific quality assurance(QA) of stereotactic radiosurgery(SRS) volumetric modulated arc therapy (VMAT), predicted dose(D-Log) was reconstructed using trajectory logfile information to compare with measurement dose(Dm). Plan optimization and beam parameters were refined to improve target dose conformity and dose agreement with D-Log analysis.
Methods: Trajectory logfiles were created during SRS-VMAT implemented in a HyperArc solution for multiple brain metastasis. All machine operation parameters are analyzed and recreated as a DICOM-RT file to recalculate the dose in a treatment planning system. To evaluate dose agreements among planned dose(D-Plan), D-Log, and Dm, a custom-made brain phantom was developed to insert films in the sagittal, coronal, and axial plane and an ion chamber at different points of interest. Isodose surfaces were used to evaluate volumetric target dose conformity and bridge doses between D-Log and D-Plan. All multi-leaf collimator (MLC) segments of D-Plan and D-Log at each gantry angle were overlapped at the beam’s eye view with leaf position errors to discern clinically significant positional errors for target and critical organs. Monitor units (MU) and relative weights for MLC segments were adjusted based on segment MU and MLC leaf errors.
Results: The D-Log showed 6% dose difference as compared with D-Plan, however, it enabled to separate dosimetric errors contributed by machine performance from composite errors. A comparison of MLC segments of D-Log and D-Plan at the beam’s eye view was useful to discern critical MLC leaf errors in small field dosimetry. Refining plan optimization parameters and segment MU improved a 5% dose agreement.
Conclusion: Trajectory logfiles effectively detected specific error components which are critical to SRS VMAT dosimetry. Dose agreement was improved by identifying the adjustable plan and beam parameters through a comparison of D-Log and Dm in SRS VMAT.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by Varian Medical Systems (Palo Alto, CA, USA).