Evaluation of image-guided positioning for frameless intracranial radiosurgery.

Abstract

The standard for target alignment and immobilization in intracranial radiosurgery is frame-based alignment and rigid immobilization using a stereotactic head ring. Recent improvements in image-guidance systems have introduced the possibility of image-guided radiosurgery with nonrigid immobilization. We present data on the alignment accuracy and patient stability of a frameless image-guided system. Isocenter alignment errors were measured for in vitro studies in an anthropomorphic phantom for both frame-based stereotactic and frameless image-guided alignment. Subsequently, in vivo studies assessed differences between frame-based and image-guided alignment in patients who underwent frame-based intracranial radiosurgery. Finally, intratreatment target stability was determined by image-guided alignment performed before and after image-guided mask immobilized radiosurgery. In vitro hidden target localization errors were comparable for the framed (0.7 +/- 0.5 mm) and image-guided (0.6 +/- 0.2 mm) techniques. The in vivo differences in alignment were 0.9 +/- 0.5 mm (anteroposterior), -0.2 +/- 0.4 mm (superoinferior), and 0.3 +/- 0.5 mm (lateral). For in vivo stability tests, the mean distance differed between the pre- and post-treatment positions with mask-immobilized radiosurgery by 0.5 +/- 0.3 mm. Frame-based and image-guided alignment accuracy in vitro was comparable for the system tested. In vivo tests showed a consistent trend in the difference of alignment in the anteroposterior direction, possibly due to torque to the ring and mounting system with frame-based localization. The mask system as used appeared adequate for patient immobilization.