Commissioning and clinical implementation of Mobius3D and MobiusFX: Experience on multiple linear accelerators

•Vendor-provided and measured beam model parameters are in a reasonable agreement.•Mobius3D DLG optimization results are affected by the selection of VMAT plans.•Excluding small-field VMAT plans from DLG optimization results in an unbiased DLG correction factor.•Small-field VMAT plans can be effectively classified by an alternative method proposed in this work. To provide practical guidelines for Mobius3D commissioning based on experiences of commissioning/clinical implementation of Mobius3D and MobiusFX as patient-specific quality assurance tools on multiple linear accelerators. The vendor-suggested Mobius3D commissioning procedures, including beam model adjustment and dosimetric leaf gap (DLG) optimization, were performed for 6 MV X-ray beams of six Elekta linear accelerators. For the beam model adjustment, beam data, such as the percentage depth dose, off-axis ratio (OAR), and output factor (OF), were measured using a water phantom and compared to the vendor-provided reference values. DLG optimization was performed to determine an optimal DLG correction factor to minimize the mean difference between Mobius3D-calculated and measured doses for multiple volumetric modulated arc therapy (VMAT) plans. Small-field VMAT plans, in which Mobius3D has dose calculate uncertainties, were initially included in the DLG optimization, but excluded later. The measured beam data were consistent across the six linear accelerators. Relatively large differences between the reference and measured values were observed for the OAR at large off-axis distances (>5 cm) and for the OF for small fields (