Program Information
Motion Tracking with the EPID and Its Dependence On Noise Versus Resolution
J Rottmann1*, Y Hu1 , M Myronakis1 , A Wang2 , P Baturin2 , R Fueglistaller3 , P Huber3 , D Morf3 , J Star-Lack2 , R Berbeco1 , (1) Brigham and Woman's Hospital, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA,(2) Varian Medical Systems, Palo Alto, CA,(3) Varian Medical Systems, Baden, Dattwil, Switzerland
Presentations
MO-L-GePD-J(A)-3 (Monday, July 31, 2017) 1:15 PM - 1:45 PM Room: Joint Imaging-Therapy ePoster Lounge - A
Purpose: Segmented scintillators for electronic portal imaging devices (EPID) enable substantial image contrast enhancements yet may feature lower resolution at high spatial frequencies inherent to their pixilation. In this study we investigate this trade-off in the context of real-time motion estimation for fiducial markers and soft-tissues – a valuable tool for online treatment delivery assurance and verification.
Methods: A 12mm thick pixelated scintillating LKH5 glass array (42.4x42.4cm²) featuring 1.51mm pixel pitch was mounted on the active matrix flat panel imager (AMPFI) of a Varian AS-1200 EPID, replacing the standard GdO2S2:Tb (GOS) screen. Fiducial gold markers typically used in the clinic as well as a 3d printed lung tumor model were embedded into a dynamic anthropomorphic thorax phantom programmed to exert a sinusoidal breathing motion of 10 mm peak-to-peak. EPID images were acquired in cine mode (7.5fps) using a 6MV photon beam in AP setup (0.52mGy/frame). The motion of the fiducial gold markers was partially occluded by bony anatomy (vertebrae). Tracking performance was assessed with an in-house developed template matching algorithm and errors estimated by fitting against a sinusoidal trajectory.
Results: While the LKH5 scintillator yielded about five times higher contrast-to-noise ratio (CNR), the MTF50 decreased by more than 50% compared to GOS. Pixilation artifacts decreased the visibility of fiducial gold markers, particularly for diameters smaller than 1mm. Visibility of the soft tissue tumor model was less affected and contrast improved. A gold fiducial with 1mm diameter could be consistently tracked with submillimeter accuracy using GOS yielding (0.3±0.2)mm average tracking error, yet LKH5 lost track intermittently resulting in an average error of (1.5±2.1)mm.
Conclusion: For the investigated scintillation materials, the lower noise levels provided by the segmented LKH5 material was not able to provide additional tracking performance as the image resolution became a limiting factor particularly for small fiducial gold markers.
Funding Support, Disclosures, and Conflict of Interest: The project described was supported, in part, by Award Number R01CA188446-01 from the National Cancer Institute. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Contact Email: