Abstract:Abstract Objective: To analyze the influence of O-arm computer-assisted navigation technology on the accuracy of spinal pedicle screw placement. Methods: Retrospective analysis of clinical data of who underwent pedicle screw placementby O-arm or free-hand in our hospital from January 2017 to September 2018. We observed pedicle screw placement and assessed the safety of pedicle screw placement. According to NEO classification, the pedicle screw placement was divided into 4 types: Grade 0: No perforation of the pedicle cortex; Grade 1: Perforations<2mm, or<50% of screw diameter outside the pedicle cortex; Grade 2: Perforations between ≥2mm and <4mm, or ﹥50% of screw diameter outside the pedicle cortex; Grade 3: Perforation of pedicle cortex, the screw is in the hole of the spinal canal or vertebral artery. The grade 0 is a precise screw placement and grade 1 is acceptable for screw placement. Grade 0 and 1 are not required to be fixed, but grade 2 and 3 are poor placement and need to be re-adjusted. Results: Totally 1459 pedicle screws were implanted in 233 patients by O-arm. According to NEO classification, The successful placement rate of pedicle screw: cervical(97.75%),thoracic(98.13%), lumbar(98.00%) and sacrum (98.00%).The average screws placement rate is 98.0%.In the free-hand group, 1742 pedicle screws were implanted. According to NEO classification, the successful implantation rates of type 0 and type 1 pedicle screws were 88.48% for cervical vertebrae, 94.84% for thoracic vertebrae, 90.57% for lumbar vertebrae, 93.10% for sacrum and 91.85% for sacrum. There was a significant difference in the classification of screw placement accuracy between the navigation group and the free-hands group.All cases successfully completed the operation, no neurovascular complications related to pedicle screw insertion were noted, the safety rate of O-arm group was 100%. However, in the free-hand group, complications such as vascular and nerve injury occurred in 4 cases, and the safety rate was 98.83%. All patients were follo