Collision Risk Mitigation of Varian TrueBeam Linear Accelerator With Supplemental Live-View Cameras.

Abstract

BACKGROUND\nNoncoplanar radiation therapy techniques such as 4π have potential dosimetric advantages but introduce complexities in treatment delivery that increase the risk for collision. Direct or remote visual confirmation of clearance is a safeguard against collisions of the gantry, couch, and patient. With our institution s Varian TrueBeam system, we identified configurations that cannot be visualized with the included closed-circuit television cameras. At our practice, electronic, portal imaging device (EPID) collision risk also exists because of the routine deployment to capture exit-dose images for treatment quality assurance. We propose a simple, cost-effective solution using network cameras to help eliminate blind spots that permits safe, noncoplanar arrangements with an EPID-acquired exit dose.\n\n\nMETHODS AND MATERIALS\nTwo Panasonic cameras were installed overhead while a third Panasonic camera was mounted onto the pedestal to monitor the couch undersurface. Live views from each camera were accessed with a web-based client. The EPID and gantry were visually assessed at 52 couch and gantry rotational angle configurations at 6 couch translational positions. Visibility was compared for the standard and supplemental camera setups at each configuration (χ2 test).\n\n\nRESULTS\nOf the 294 assessable couch-gantry configurations, the standard camera setup had limited visibility of either gantry or EPID for 146 configurations compared with 72 configurations with additional cameras (51% blind-spot reduction; P < .01). An 87% blind-spot reduction was observed for our laterally centered, cranial-based, couch translational position (P < .01).\n\n\nCONCLUSIONS\nThe supplemental cameras were simple, effective additions for collision detection, especially for noncoplanar radiation therapy with EPID-acquired, exit-dose imaging. Over half of the assessable noncoplanar configurations had blind spots using standard cameras, which was reduced to <25% with additional cameras. In practice, there were almost no blind spots for patients with brain tumors who were treated with our templated beam arrangements. Using live-view camera feeds, vault re-entry to visually confirm clearance was reduced approximately 10-fold, which increased the treatment efficiency. In the most recent 12 months, no collision or near-collision events have been reported.