ABSTRACT
STUDY DESIGN: The study involved the development and accuracy testing of an intra-operative navigation system. OBJECTIVES: This study was undertaken to develop a navigation system using a robot arm-type three-dimensional digitizer. And, to apply the developed system to pedicle screw insertion, and to evaluate its accuracy. SUMMARY OF LITERATURE REVIEW: To the best of our knowledge, no navigation system has been developed using a robot armtype three-dimensional digitizer. MATERIALS AND METHODS: We have developed a navigator using a three-dimensional digitizer (Microscribe 3-D G2, Immersion, USA) supported by a personal computer. Four types of patient-to-image registration techniques were implemented. During navigation, the central axis of the robot arm's stylus and arm extension can be displayed over multi-planar and three-dimensional images, which are reconstructed from axial CT scan images. Registration errors and target localization errors of the navigation system were evaluated using a phantom made from a plastic lumbo-sacral bone model. The accuracy of pedicle screw insertion was also evaluated by placing 18 pedicle screws in such bone models. RESULTS: The registration error was 0.78 +/- 0.27 mm at fiducial registration and 0.76 +/- 0.24 mm at hybrid registration, and the target localization error was 1.34 +/- 0.32 mm at fiducial registration and 1.28 +/- 0.29 mm at hybrid registration. Of the 18 screws placed in the plastic bone models, one (6%) screw breached the pedicle wall. CONCLUSIONS: We have developed a robot arm-type three-dimensional digitizer-based navigation system for pedicle screw insertion, and found that its accuracy is equal or slightly better than that of optical tracker-based navigators.