Room: Exhibit Hall
Purpose: A dose volume histogram (DVH) based analysis with three-dimensional detector is convenient for the patient specific quality assurance on VMAT (VMAT-QA). For the analysis, the DVH and gamma passing-rate were important factor. However, there are several differences in each analyzing algorithm among software and it is unclear how we decide suitable criteria for that. The purpose of this study is to make sure the analytical difference among the software and to establish suitable criteria for VMAT QA.
Methods: Two three-dimensional detectors and their DVH based analysis were evaluated: 3DVH with ArcCHECK (SunNuclear) and Anatomy with Delta4 (ScandiDos). As fundamental experiments, several demonstrative plans were created to water-equivalent phantom. As clinical experiments, 10 patients with prostate cancer were chosen for VMAT-QA. With two 3D detectors, we evaluated both the dose at several region-of-interests (ROIs) and 2D gamma passing-rate for their dose distribution. Then we computed DVH on these plans for comparing with dose criteria in our hospital and for calculating 3D gamma passing-rate.
Results: The measured absolute dose errors at all ROIs were satisfactory matched with planed dose. However, the 2D gamma passing-rates in dosimetric verifications with Delta4 were higher than that with ArcCHECK in both of fundamental and clinical experiments. In addition, 3D gamma passing-rates at all evaluated regions with 3DVH were higher that with Anatomy. Therefore, the volumetric dosimetric verification by 3DVH with ArcCHECK was sometimes overestimated for satisfying clinical dose criteria than that by Anatomy with Delta4.
Conclusion: We evaluated dosimetoric verification with two 3D detectors to make sure analytical difference due to their analysis algorithm and geometry of detectors. We could make sure that 3DVH with ArcCHECK is easily able to satisfy the dose criteria in our hospital. The dose criteria for DVH based VMAT-QA should be defined individually for 3D detectors with its analysis software.