Session Item

Our centre has recently implemented Ethos Therapy system (Varian, Palo Alto) that is capable of delivering online-adaptive radiotherapy based on CBCT images. For online adaptive radiotherapy an efficient and reliable quality assurance (QA) process is required. Phantom-based patient specific-QA (PSQA) is not suitable for on-line adaptive radiotherapy (ART). Varian provides Mobius3D with Ethos Therapy, which incorporates a beam model independent from the planning system as a tool for PSQA. As the products are supplied by the same vendor, the question of independent PSQA is mitigated by complementing the QA process with phantom-based QA on a monthly basis to assess the stability of machine performance. DoseCHECK (Sun Nuclear Corporation) is a third-party solution which is capable of performing PSQA for online ART. Mobius3D and DoseCHECK are both capable of using trajectory log files to calculate dose distributions. Treatment log files, together with the pre-delivery check of the Ethos Therapy TPS dose calculation provide effective alternative to phantom‐based PSQA measurements. The objective of this study was to establish the use of DoseCHECK for online adaptive radiotherapy treatment and compare its sensitivity with Mobius3D to intentional errors introduced into the treatment plans generated by Ethos therapy TPS.

7-field IMRT prostate plans were produced by Ethos TPS which uses Acuros v15.6 algorithm. The exported plans from the Ethos TPS were edited to introduce intentional errors: MLC bank shifts; collimator rotations; MU changes for a number of control points around the centre of each beam; and a random positional error for all MLC leaves used for modulation. These edited plans were dose distribution calculated by the Ethos TPS.

Mobius3D and DoseCHECK are sensitive to 0.5mm MLC bank shifts with an average reduction of 5.5% and 1% respectively for 2%/2mm gamma criteria, threshold 10%. Mobius3D was more sensitive to random MLC positional errors. The two systems have similar sensitivity to collimator rotation with a 1% reduction in gamma passing rate per degree of collimator shift between 0⁰ and 5⁰. DoseCHECK and Mobius3D were found to be sensitive to MU change. Calculation times of the two systems were found to be comparable.

Both Mobius3D and DoseCHECK have shown good sensitivity to a multitude of intentionally introduced errors and both have the capability to act as the primary PSQA system for Ethos ART. There are differences between the sensitivities of the two systems to a given plan error type, with Mobius3D being more sensitive to MLC positional error. However, through audit and analysis of results comparing them to phantom measurement, users of either system would be able to set their PSQA gamma criteria and pass/fail tolerances so that relevant plan errors would be detected.