Room: AAPM ePoster Library
Purpose: to clinically implement IMRT QA tolerance and action limits in the manner recommended by AAPM Task Group 218.
Methods: A total of 80 patient-specific VMAT QA plans were analyzed from planar dose distributions measured with an ArcCHECK. The plans consisted of 20 initial prostate, 20 prostate boosts, 20 initial head and neck and 20 head and neck boosts. The global gamma analysis criteria of 3%/3mm was compared with the new standard, 3%/2mm, using a 10% dose threshold. The resulting measurements were used to determine process-based tolerance and action limits for the different plan types. Finally, random and systematic errors were introduced to determine the sensitivity of the new passing criteria to common failure modes in the IMRT QA process.
Results: Bulk analysis of all 80 plans showed an average passing rate reduction of 1.3% (99.3% to 98.0%). From these plans, our institutional tolerance and action limits for the new standard were calculated to be 93% and 92%, respectively. The passing rates showed no clear trend when comparing prostate to head and neck plans. However, for both treatment sites, boost plans showed a significant reduction in passing rates when compared to initial plans. The process-based action limits showed the increased variability between different plan types. They were 94% for the Initial Prostate and Initial H&N and 90% for the Prostate Boost and H&N Boost. The new criteria showed greater sensitivity when an absolute dose calibration error of 1.5% was introduced. This artificial systematic error reduced passing rates of individual plans by up to 10%, and 2% on average.
Conclusion: Transitioning to the TG218 recommendation for gamma passing criteria shows an overall reduction in passing rate and increased variability across different plan types. Process-based tolerance and action limits can be coupled with root-cause analysis to improve IMRT system performance.