Room: AAPM ePoster Library
Purpose: SRS head frames obscure parts of the face typically used for surface tracking during SRS. This work investigated the feasibility of surface tracking during framed treatments using Varian OSMS.
Methods: An SRS head frame was placed on a Rando phantom with three implanted BBs at various locations in the brain. Treatment plans were created with isocenters centered on each BB. The phantom was initially localized using CBCT with the couch at 0°. A new reference surface was acquired with OSMS after CBCT-guided shifts. A region-of-interest was defined in OSMS that avoided the posts of the head frame that obscured parts of the face. From 0°, the couch was rotated to +/-45° and +/-90°. At each couch angle, the OSMS offsets were recorded and an AP MV image was acquired using a 2x2 cm² field. The RIT software Winston-Lutz module was used to determine lateral and longitudinal offsets of the BB from the MV image. The OSMS-indicated offsets were compared to the RIT-determined offsets. Additionally, the effect of having one of the OSMS cameras blocked during tracking was evaluated (i.e., gantry at +/-45°, depending on couch angle). OSMS offsets before and after a camera was blocked were compared at each couch rotation.
Results: The OSMS-indicated offsets were all within 0.75mm of the RIT-determined offsets of the implanted BBs for all three isocenters. Vertical offsets were not evaluated radiographically, but the largest OSMS-indicated vertical offset was +/-0.1mm. The largest change in OSMS-indicated magnitude offset (i.e., the vector sum of the translations) when a camera was blocked was 0.3mm and the largest change in any rotational offset was 0.3°.
Conclusion: OSMS demonstrated sub-millimeter agreement with Winston-Lutz analysis of implanted BBs at three locations within a head phantom. This accuracy supports the feasibility of tracking framed SRS patients with this system during treatment.