Assessment of the Dosimetric Performance of the Mobius3D against Portal Dose Measurements in Patient-specific Quality Assurance.

Abstract

Aim: The Mobius3D software addresses limitations lacking in measurement-based methods in patient-specific quality assurance (QA). The objective of this study was to validate its dosimetric performance against conventionally used portal dose measurements using gamma analysis and confidence limits.

Materials and methods: A total of 240 patient-specific QA plans for the Varian Halcyon linear accelerator were collected. The Mobius3D software was commissioned through beam data and plan verification. All plans underwent QA through the electronic portal imaging device, coupled with the Portal Dosimetry software, and the Mobius3D. Data were assessed using >95% gamma pass. Portal measurements were evaluated using 3%/2 mm and 3%/3 mm criteria, whereas Mobius3D was analyzed at 3%/3 mm and 5%/3 mm, at the 10% threshold.

Results: Mobius 5%/3 mm mean gamma passes were 99.89% for volumetric-modulated arc therapy (VMAT) and 99.31% for intensity-modulated radiotherapy (IMRT), and correspondingly, the data for portal 3%/2 mm were 99.99% and 99.96%. The Mobius3D at 5%/3 mm can perform like Portal 3%/2 mm for VMAT plans at 0.1% difference, especially for head/neck and pelvic/abdominal cases. In IMRT-based treatments, at 0.7% difference in Mobius3D 5%/3 mm and Portal 3%/2 mm, the performance and error identification in IMRT plans should be applied more carefully due to the amount of failed plans, particularly the chest region. The confidence limits for VMAT plans for Portal 3%/2 mm and Mobius 5%/3 mm are 99.93% and 99.42%, respectively, while for IMRT plans are 99.69% and 97.43%, respectively.

Conclusions: At a 5%/3 mm criterion, the Mobius3D may yield percentage gamma pass rates like measurements obtained by Portal Dosimetry 3%/3 mm and Portal Dosimetry 3%/2 mm. As the software is largely dependent on commissioned data, rigorous commissioning and a comprehensive QA program should be implemented.