Program Information
The Effect of Planning CT Slice Thickness On Evaluation of SRS Target Localization Accuracy
J Briceno*, H Li , Y Huang , N Wen , I Chetty , Henry Ford Health System, Detroit, MI
Presentations
SU-E-J-30 (Sunday, July 12, 2015) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: To evaluate the effect of planning CT slice thickness on SRS target localization accuracy on Varian Edge SRS treatment platform.
Methods: A metal BB of diameter 5 mm was inserted in a Rando head phantom. Planning CTs were scanned with 0.8, 2 and 3 mm slice thickness respectively. The center of BB was setup as isocenter and tattoos were marked on phantom surface. The phantom was initially aligned using laser and tattoos, then a CBCT was acquired and 6-degree of freedom (DOF) image registration was performed based on bony anatomy. The 6D couch correction was applied. A verification CBCT was taken to evaluate couch corrections. Three orthogonal pairs of images (MV-MV, KV-KV and MV-KV), were also taken to verify the position of the phantom using a 2D-3D registration algorithm. Winston-Lutz tests were taken at12 combinations of gantry, collimator and couch angles using 1x1cm MLC aperture and 10mm circular cone.
Results: The vector lengths of the residual error measured with verification CBCT were 0.3, 0.2 and 0.2 mm for the planning CT of 0.8, 2 and 3 mm slice thicknesses respectively. Residual setup error vector lengths for 0.8, 2, and 3 mm CT measured by orthogonal pair were as following: a) MV-MV: 0.2, 0.5, 0.5mm; b) KV-KV: 0.2, 0.3, 0.3mm; c)MV-KV: 0.1, 0.6, 0.3mm respectively. The vector lengths for the Winston-Lutz tests were: a) MLC: 0.55, 0.32, and 0.2mm; b)CONE: 0.16, 0.17, and 0.29mm for slice thickness of 0.8, 2, and 3mm respectively.
Conclusion: The residual setup error in this study measured by the 2D-3D algorithm was comparable to that of the 3D-3D algorithm. The planning CT slice thickness in the range of 0.8 to 2 mm has no significant impact on the evaluation of the residual setup error and the result of Winston-Lutz test.
Contact Email: