Wednesday, September 26, 2018 7:35 AM–9:00 AM ePosters

Abstract

BACKGROUND CONTEXT Several studies have shown that the use of a combination of intraoperative CT images and navigation, significantly improves pedicle screw placement accuracy. A new concept of surgical navigation system using augmented reality (AR) with intraoperative 3D cone beam CT (CBCT) was developed specifically for spinal surgery. PURPOSE To evaluate accuracy, efficiency, and safety of an AR-based surgical navigation system for pedicle screw placement. STUDY DESIGN/SETTING Prospective study. PATIENT SAMPLE Twenty patients. OUTCOME MEASURES Pedicle screw placement accuracy, procedural time, patients and occupational radiation doses. METHODS Twenty patients (1:1 male:female, 30±19 years) eligible for spine surgery with pedicle screw placement (13 scoliosis deformity cases and 7 other surgical indications) were prospectively enrolled and treated using surgical navigation with AR (Philips Healthcare, Best, the Netherlands). A total of 253 screws were placed in the patients using AR. The navigation system consists of optical video cameras integrated in a ceiling-mounted C-arm in a hybrid operating room (OR). An intraoperative CBCT was acquired from the C-arm to plan an optimal path for screw placement. The screws were navigated thanks to the planned paths augmented to the reality scene shown by the video cameras. An additional CBCT was performed to evaluate the accuracy of screw placement by independent reviewers using the Gertzbein scale. Total navigation time for screw placement, patient radiation dose as well as occupational staff dose was measured. RESULTS In average 13±6 screws per patient were placed with navigation. The accuracy of pedicle screw placement was 94% yielding 15/253 screws rated as grade 2 (2–4 mm pedicle breach) from which 13 were lateral breaches. Thirteen of these grade 2 screws had a diameter larger than the pedicle isthmus size where they were placed. The average navigation time was 5.2±4.1 minutes with a median value of 4 minutes from making the bone entry burr-hole to the final screw placement. Navigation time corresponded to an average of 16.8% of the total procedure time. The average patient radiation dose was 31±13 Gy cm2 (median 32, range: 14–51) with an average of 4 CBCT scans per procedure. During intraoperative 3D acquisitions, the OR staff was behind a leaded shield in the corner of the OR and did not wear apron lead protection during surgery. The average cumulative occupational dose of all staff member in the OR was 0.8±1.1 µSv (median 0.3, range: 0.0–3.2). CONCLUSIONS Surgical navigation using AR with intraoperative CBCT demonstrated clinical accuracy as high as 94% of pedicle screw placement with no related clinical complications. The average required time of navigation per screw placement was as short as 5.2 minutes. The use of navigation was a short fraction of the total procedure time. The CBCT intraoperative imaging available on the AR-based navigation system demonstrated negligible cumulative staff dose.