Technical note: Clinical real-time multi-target, 3D-motion tracking radiation delivery on a magnetic resonance linear accelerator

BACKGROUND: 
The purpose of this study was to demonstrate the feasibility of multi-target 3D-motion tracking for use in magnetic resonance-imaging (MRI)-guided, stereotactic body radiotherapy for upper abdominal targets. This paper focuses on the clinical testing of a real-time 3D-motion tracking process for both clinical targets and organs-at risk-for automatic beam gating in MRI-guided radiation delivery (MRgRT).
Patient and methods:
A pilot study was performed using a commissioned workflow based on an in-house dual-target phantom. A patient with primary liver cancer in segment 5 with adjacent bowel was selected for dual-plane tracking, two target tracking, and OAR dose-accumulation for auto-beam gating. Data from all five delivery fractions were reviewed and presented, including cine MRI settings, target tracking settings, respiratory motion management, and delivery details.
Results: The initial delivery time for the first fraction was 35 minutes, substantially longer than the expected duration of 11.8 minutes, which was determined based on a 50% delivery time using the calculated monitor units (MU). Subsequent treatments were consistently faster, with the last fraction being 9 minutes. For dual-plane tracking, the clinical team defined tracking parameters based on the motion and location of the secondary tracking structure (critical organ: large bowel) in relation to the tumor, using a 6 mm boundary expansion. This was compared to a tighter 3 mm boundary expansion for the primary target. 
Conclusions: We demonstrated the clinical feasibility of dual-plane tracking in MRgRT which enables multi-target, 3D-motion tracking of both targets and organs at risk. This approach can provide an additional motion‑monitoring capability beyond conventional single‑plane tracking while maintaining effective target gating and treatment time. This initial institutional implementation demonstrates the technical feasibility of dual‑plane tracking and its potential to incorporate motion monitoring of selected critical structures in MRgRT.&#xD.