SU-FF-J-110: Geometrical Verification of Real-Time Tumor Tracking Using Electronic Portal Imaging Device (EPID)

Abstract

Purpose: We have developed a novel method to verify the geometrical accuracy of real‐time tumor tracking. An Electronic Portal Imaging Device(EPID) is used as a virtual phantom system. Methods and Materials: This virtual phantom system uses virtual target motion in the absence of a moving phantom and fluoroscopic mode of EPID for monitoring the motion of the tracking aperture. The expected apertures are designed from real‐time detection of tumor motion, and the delivered apertures are monitored from the EPID system. These two apertures are compared for geometrical accuracy of a real‐time tracking technique. The virtual phantom system was tested with several tracking patterns with irregular motion. Target motion was designed for an oval‐shaped tumor with 40 mm and 55 mm in diameters for four cases: (1) stationary; (2) rigid‐body two‐dimensional (2‐D) displacement with a period variation from 3 to 10 s and 20‐mm peak‐to‐peak distance in the superior‐inferior direction and 10 mm in the left‐right direction; (3) 2‐D displacement with deformation, and (4) 2‐D displacement with both deformation and rotation. In the continuous acquisition mode of EPID,MLC aperture images were acquired at ∼8 Hz. Root‐mean‐square (RMS) deviation between the designed and delivered aperture motions were calculated for each case. Results: The RMS deviation is 0.005 mm for the stationary case, 0.5 mm for the 2‐D displacement, 0.7 mm for the 2‐D displacement with deformation, and 0.7 mm for the 2‐D displacement combined with both deformation and rotation. Conclusion: The EPID‐based virtual system can measure geometrical accuracy of real‐time tumor tracking without a moving phantom. Besides, it can be used even for complex target motion (i.e., 2‐D displacement combined with deformation and rotation), which is not possible with any existing moving phantom. Therefore, this EPID‐based virtual system shows great potential as a quality‐assurance process to verify real‐time tumor tracking.