Predicting Delivery Error Using a DICOM-RT Plan for Volumetric Modulated Arc Therapy

The purpose of this study was to investigate the prediction of mechanical error using DICOM-RT plan parameters for volumetric modulated arc therapy (VMAT). We created plans for gantry rotation arcs of 360° and 180° (full-arc and half-arc VMAT) for six maxillary sinus cancer cases using a Monaco treatment planning system, and delivered the doses with a linear accelerator. We calculated DICOM-RT plan parameters, including gantry, multileaf collimator (MLC) positions and Monitor Units (MU). We compared plans with regard to gantry angle per MU (degrees/MU) and MLC travel per MU (mm/MU) for each segment. Calculated gantry angle/MLC position speeds and errors were evaluated by comparison with the log file. On average, the half-arc VMAT plan resulted in 47% and 35% fewer degrees/MU and mm/MU than the full-arc VMAT plan, respectively. The root mean square (r.m.s.) gantry and MLC speeds showed a linear relationship with calculated degrees/MU and mm/MU, with coefficients of determination (R2) of 0.86 and 0.72, respectively. The r.m.s. gantry angle and MLC position errors showed a linear relationship with calculated degrees/MU and mm/MU with R2 of 0.63 and 0.76, respectively. Deviations from plan parameters were related to mechanical error for VMAT, and provided quantitative information without the need for VMAT delivery. These parameters can be used in the selection of treatment planning.