Program Information
Comparison of BrainLab ExacTrac and CBCT Position Accuracy for Spine SBRT On a Truebeam STX
J Ha*, J Ye , A Shiu , E Chang , Keck School of Medicine of USC, Los Angeles, CA
Presentations
TU-RPM-GePD-J(B)-1 (Tuesday, August 1, 2017) 3:45 PM - 4:15 PM Room: Joint Imaging-Therapy ePoster Lounge - B
Purpose: To compare the on-line 6 degree-of-freedom image registration of Truebeam STX CBCT and BrainLab ExacTrac imaging system for spine SBRT.
Methods: Winton-Lutz test was conducted to confirm the center of the two imaging system coincided with the linac isocenter. A BrainLab pelvis phantom was used to verify the two registration algorithms. The phantom was first aligned using CBCT. Two orthogonal ExacTrac images were taken, and the image registration software was run 50X to determine the mean and standard deviation of the shift coordinates. Further, the 6D perfect pitch couch was displaced and rotated by about 1mm or 1deg to see if the registration software could detect the change accurately. The ExacTrac shift data for 5 SBRT patients are also analyzed in this study.
Results: The Winton-Lutz test confirms the centers of the two imaging systems stay within the 0.4mm radius of the linac isocenter. The mean position and orientation of the phantom as determined by ExacTrac were slightly different from CBCT, by about 0.15 to 0.3 mm and 0.2 to 0.5 deg, respectively. Because the registration software generates different shifts each time, there were instances in the phantom study where the ExacTrac position and rotation shifts were 0.4 mm and 0.55 deg, even after using CBCT to align the phantom. Both CBCT and ExacTrac detect the 1mm or 1 deg shift accurately to about 0.1mm or 0.1 deg accuracy. The data from the 5 SBRT patients are very similar to the phantom study.
Conclusion: If the setup tolerance is 1 mm for displacement and 1 deg for rotation, our data shows that ExacTrac can meet the challenge. If the tolerance is 0.5mm and 0.5 deg, caution should be taken to verify ExacTrac data against CBCT and access the mean and spread in ExacTrac data for the patient-specific DRRs.
Contact Email: