Purpose: An automated, comprehensive in-vivo system to detect changes in patient anatomy and machine output was developed using a novel analysis of in-vivo EPID images for every fraction of treatment on a Varian Halcyon. The in-vivo approach was used to identify errors that go undetected by routine quality assurance (QA) to compliment daily machine performance check (MPC), with minimal additional physicist workload.
Methods: Images for all fractions treated on Halcyon were automatically downloaded and analyzed at the end of the treatment day. For image analysis, we select a high-dose region that has been shown to be a predictive metric for changes in the PTV mean dose (PTVµ). Flags are raised for: (Type-A) patients whose PTVµ exceeds 10%, to protect against large errors, and (Type-B) patients with three consecutive fractions higher than 3%, to protect against systematic trends. A detailed report for flagged patients is e-mailed to a physicist for further investigation. Furthermore, daily averages and standard deviations are uploaded to Total QA, our QA portal, in addition to MPC that are signed-off by physicist daily, ensuring comprehensive QA for Halcyon. To guide clinical implementation, beginning on 12-15-2020, a retrospective study was conducted for treatments from Nov-2018 till Dec-2020, grouping errors by treatment site.
Results: From retrospective data of 1033 patients, no Type-A errors were found and only 31 patients (3 %) had a Type-B error. Most of these deviations for Type-B were due to Head/Neck weight loss. For 3 months of use, no flags have been raised by the system that required immediate intervention, indicating the system is running nominally.
Conclusion: This system protects against errors that can occur in-vivo to provide a more comprehensive QA. This fully automated framework can easily be implemented in other centers with a Halcyon and only requires a standard desktop computer and analysis scripts.
Funding Support, Disclosures, and Conflict of Interest: Supported by grant funding from Varian Medical Systems.