Background: The objective of the study was to improve the accuracy of computer-assisted pedicle screw installation in the spine. This study evaluates the accuracy of computer-assisted pedicle screw placement with separate spinal reference clamp placement and registration on each instrumented vertebra for thoracolumbar spine fractures.
Methods: Postoperative radiographs and CT scans assessed the accuracy of pedicle screw placement in 21 adult patients on each instrumented vertebra. Screw placements were graded as good if the screws were placed in the central core of the pedicle and the cancellous portion of the body. Screw placements were graded as fair if the screws were placed slightly eccentrically, causing erosion of the pedicular cortex, and with less than a 2-mm perforation of the pedicular cortex. Screw placements were graded as poor if screws were placed eccentrically with a large portion of the screw extending outside the cortical margin of the pedicle and with more than a 2-mm perforation of the pedicular cortex.
Results: A total of 140 image-guided pedicle screws were placed in 21 patients: 78 in the thoracic and 62 in the lumbar spine. Of the 140 pedicle screw placements, 96.4% (135/140) were categorized as good; 3.6% (5/140), fair; and 0% were poor. All 5 fair placement screws were placed in the thoracic spine without any mobility.
Conclusion: Separate registration increases accuracy of screw placement in thoracolumbar pedicle instrumentation. Separate spinal reference clamp placement in the instrumented vertebra provides real-time virtual imaging that decreases the possibility of downward displacement during manual installation of the screw.