Six-Dimensional Image Guidance For Spinal Noninvasive Srs/Sbrt: Comparison Between Exactrac X-Ray 6d With Kv Cone-Beam Ct

This study was to investigate the 6D image guidance with ExacTrac X-ray 6D and cone-beam CT (CBCT) for patients under treatments of spinal SRS/SBRT on a Novalis Tx system. The Novalis Tx system is equipped with 6D robotic couch, ExacTrac X-ray 6D, and KV CBCT. In this work, prospective phantom and retrospective patient studies were performed to investigate localization accuracy for image-guided SRS/SBRT with ExacTrac X-ray 6D and CBCT. In the phantom studies, a pelvis phantom was positioned on the couch with a predefined 6D shift including pitch, roll, and raw. The phantom was then imaged with ExacTrac X-ray 6D to yield a 6D shift. After this, the phantom was scanned with CBCT, and matched with planning CT using CBCT 6D image registration (CBCT 6D6D match) to yield another 6D shift. The matching results were used to test the accuracy of ExacTrac X-ray 6D and CBCT 6D6D match for correcting translational and rotational setup errors. In patient studies, 5 spine SBRT patients were retrospectively studied. The patients were initially positioned in customized immobilization cradles. Their setup positions were then corrected by using ExacTrac X-ray 6D and 6D robotic couch. After this, the patients were scanned with CBCT, and positions were verified or adjusted based on CBCT. The setup discrepancies were quantitatively analyzed. The phantom experiments showed that root-mean-square (RMS) of residual setup errors in translations and rotations were <0.8 mm and 0.2o by CBCT 6D6D match. In contrast, the residual errors of ExacTrac X-ray 6D in translations and rotations were, respectively, <1.3 mm and 0.6o. Translational discrepancies in RMS between ExacTrac X-ray 6D and CBCT 6D6D match were 0.66 mm laterally, 0.75 mm longitudinally, and 0.70 mm vertically. Rotational discrepancies in RMS between ExacTrac X-ray 6D and CBCT 6D6D match were 0.17o, 0.43o, and 0.62o in pitch, roll, and yaw, respectively. In the retrospective patient studies, translational discrepancies in RMS between ExacTrac X-ray 6D and CBCT 6D6D match were 1.06 mm laterally, 1.43 mm longitudinally, and 1.43 mm vertically. The corresponding rotational discrepancies in RMS were 1.20o, 0.64o, and 1.66o in pitch, roll, and yaw. If positioning was corrected using CBCT 3D3D match, translational discrepancies in RMS were 0.45 mm laterally, 1.26 mm longitudinally, and 0.45 mm vertically, and rotational displacements in RMS were 1.09o, 0.44o, and 0.74o in pitch, roll, and yaw, as compared with CBCT 6D6D match. There are small, but maybe clinically-considerable discrepancies between ExacTrac X-ray 6D and CBCT 6D6D match. The CBCT 6D6D match can verify positioning in 6D, and could effectively minimize setup errors when combining with a 6D couch. Precaution should be taken when only ExacTrac 6D is used.