Purpose: To analyze treatment planning system calculation accuracy relative to patient-specific quality assurance (PSQA) measurements among two TomoTherapy treatment delivery systems and two treatment planning systems and evaluate any dependencies observed based on machine, planning system, or treatment site.
Methods: Data were extracted from PSQA reports created from Delta4+ measurements on two TomoTherapy HDA units over more than two years, during which time treatment planning was transitioned from Accuray Planning Station (PS) to RaySearch RayStation (RS). Data points included the treatment site, the median of the planned-to-measured dose deviations over all detectors, and the proportions of detectors passing gamma calculation with a range of dose and distance to agreement criteria. These data were analyzed based on sorting by treatment machine, originating treatment planning system, and by treatment site.
Results: Median dose deviation histograms considering all treatment sites revealed Gaussian probability distributions with mean values of -0.5% for PS and -0.3% for RS on Machine A, and -0.2% for PS and +0.3% for RS on Machine B. Welch’s t-tests demonstrated that all these distributions were unique, indicating that despite their independent calibrations all delivery and treatment planning system combinations produced good agreement between dose prediction and delivery. Gamma pass rates pooling all treatment sites were 98-100% for 3%/3mm evaluation criteria, dropping to approximately 91-96% for 3%/2mm and 60-66% for 2%/2mm.
Conclusion: Median dose deviation and gamma analysis both suggest similar dose calculation and delivery performance for both the Accuray and RaySearch treatment planning systems on two independent delivery systems. Because no clear dependence was observed based on generalized treatment site, future work is focused on investigating correlations with other plan characteristics such as target volume, machine parameters, and plan complexity.