(Monday, 7/15/2019) 1:15 PM - 1:45 PM
Room: Exhibit Hall | Forum 4
Purpose: Patient-specific quality assurance (PSQA) results contain in addition to cases with alerts a large set of data without alerts that may also include clinically relevant information. It is the purpose of this study to develop and test a general procedure based on systematic long-term trend analysis of PSQA data of VMAT treatments to discover site-specific deviations, and to reduce the underlying systematic uncertainties.
Methods: The proposed procedure starts with an exhaustive analysis of available VMAT dose verification data per treatment site. If unexplained systematic deviations are present, several actions are required, including confirming the observed deviations with an independent dose verification system, relating the deviating QA results with plan complexity, re-planning the deviating plans using the latest beam modeling recommendations for the TPS, and comparing the updated results with the measured QA data. The procedure was tested for a comprehensive PSQA data base of prostate VMAT verifications, collected with 3D EPID-based in vivo dosimetry.
Results: The prostate VMAT PSQA data showed a gradual lower dose of about 2% in 3 years. The deviating EPID data were confirmed by measurements with a 2D detector array in a 4D phantom, and showed a correlation with plan complexity expressed as MU/cGy. Further analysis showed that the under-dosage coincided with the transition of one- to two-arc VMAT. The implementation of new beam fits in the TPS lead to a reduction in size of the observed deviations, bringing about almost all PSQA data within tolerance.
Conclusion: Long-term trend analysis and data visualization of PSQA data of VMAT treatments are able to trace site-specific deviations and should therefore be part of a routine PSQA program. The procedure to analyze in a systematic way unexplained deviations has led to an improvement of the PSQA process and a reduction of systematic dosimetric uncertainties in VMAT treatments.
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
