Premium
Sensitivity of the IQM transmission detector to errors of VMAT plans
Author(s) -
Razinskas Gary,
Wegener Sonja,
Greber Johannes,
Sauer Otto A.
Publication year - 2018
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1002/mp.13228
Subject(s) - ionization chamber , reproducibility , standard deviation , detector , quality assurance , sensitivity (control systems) , imaging phantom , signal (programming language) , mathematics , monitor unit , nuclear medicine , statistics , physics , computer science , optics , medicine , ionization , engineering , electronic engineering , ion , pathology , quantum mechanics , programming language , external quality assessment
Purpose The integral quality monitor (IQM) transmission detector is a wedge‐shaped large area ionization chamber that reports a position‐weighted dose area product for each control point of an IMRT or VMAT plan. In this study, the accuracy of the signal prediction is verified for the Synergy Agility MLC. Tolerance criteria for VMAT plan verification with the IQM were obtained from the observed sensitivity for the detection of incorrectly delivered plans. Methods The predicted IQM signal was compared to the measured signal recorded for a set of 30 VMAT plans for each beam quality of 6 and 10 MV. The system's capability to detect incorrectly delivered plans was tested by measuring altered plans containing small, random deviations. In addition, the observed deviations were related to measurements performed with a second QA phantom. Results The cumulative IQM signal per arc deviated from the respective calculation on average by −0.48% (6 MV) and +0.21% (10 MV) with a standard deviation of 1.08% in both cases, suggesting a 2% warning and 3% action threshold as plan acceptance criteria. This choice was confirmed by the optimum threshold of 2.5% obtained via receiver operating characteristic (ROC) analysis. Reproducibility of individual control points in multiply measured plans was low (on average 7% for 1SD) and thus, segment‐by‐segment comparison was impractical. A suitable criterion to resolve the angular distribution of the plan was identified by binning three to five control points as a running average. While the correlation between IQM signal deviations and gamma passing rates obtained with the ArcCHECK phantom was low for clinical plans, it was apparent for erroneous plans. Binning led to even higher sensitivity to errors. Conclusions The IQM was able to detect induced errors at least as reliable as the standard phantom and showed the potential to be used in pretreatment plan verification to ensure the correct plan transfer and delivery. However, there is no direct correlation between the IQM signal deviation and DVH metrics, so the IQM should be primarily used to screen for errors. Finer diagnostics should then be carried out using a different phantom.