Image-Guided Target Localization for Stereotactic Radiosurgery:Accuracy of 6D versus 3D Image Fusion

Purpose: This study was aimed to demonstrate the accuracy of an image-guided target localization system with 6D image fusion in a dedicated linear accelerator radiosurgery unit, and compared to that with the 3D fusion. Materials and Methods: An anthropomorphic phantom implanted with multiple 2 mm metal spheres was used for the study. The phantom was CT scanned with 2 mm and 3 mm slice thicknesses. Experiments were performed to evaluate the localization accuracy at 10 separate isocenters in various regions of cranium and body in the phantom. Isocenters were placed at the center of the implanted markers. Translational and rotational deviations were intentionally created at the initial setup for each isocenter. Following localization with 6D image fusion, a set of kV X-rays and MV portal images were taken to verify the isocenter position. 3D image fusion was also performed for comparison. Results: The accuracy of 6D fusion localization was consistently less than 1 mm for all 10 isocenters. There was no significant difference in localization accuracy (average 0.7 mm) by the thickness of CT slices between 2 mm and 3 mm. For various translational and rotational deviations in the initial setup, the accuracy of 6D fusion method was less than 1 mm, compared to that of 3D fusion where the accuracy margin could be 2-10 mm depending on the degree of initial positioning. Conclusion: The 6D image fusion method achieved the accuracy of target localization within 1 mm for radiosurgery of cranium and body sites. This can lead to a step forward to the non-invasive positioning for radiosurgery.