Related Experiment Video
Updated: Jun 8, 2025
A Spine Robotic-Assisted Navigation System for Pedicle Screw Placement
Published on: May 11, 2020
Robotic pedicle screw placement with 3D MRI registration: moving towards radiation free robotic spine surgery
Franziska C S Altorfer1, Michael J Kelly2, Fedan Avrumova2
1Department of Spine Surgery, Hospital for Special Surgery, 523 East 72nd Street, New York, NY, USA; University Spine Center Zürich, Balgrist University Hospital, University of Zürich, Zürich, Switzerland.
Background Context:
Preoperative imaging for lumbar spine surgery often includes magnetic resonance imaging (MRI) for soft tissues and computer tomography (CT) for bony detail. While CT scans expose patients to ionizing radiation, whereas MRI scans do not. Emerging MRI techniques allow CT-like 3-dimensional (3D) visualization of bony structures, potentially removing the need for ionizing radiation from CT scans.
Purpose:
This study aims to explore the accuracy of robot-assisted lumbar pedicle screw placement based on preoperative CT-like 3D MRI as the data source for robotic registration.
Study Design:
Human cadaveric study.
Methods:
CT-like 3D MRI scans of the lumbar spine were acquired in ten human cadavers. A robotic navigation platform was used to plan and navigate pedicle screw placement based on the CT-like 3D MRI. Postoperative CT scans assessed the accuracy of screw positioning compared to preoperative planning based on the Gertzbein-Robbins scale (GRS) and by direct measurement (mm).
Results:
A total of 100 lumbar pedicle screws were robotically placed in ten cadavers (L1 through L5 bilaterally) based on CT-like 3D MRI. On postoperative CT evaluation, 99.0% of the positioned screws achieved an acceptable grade on the GRS (Grade A: n=89 or Grade B: n=10), with 89.0% classified as Grade A and 10.0% as Grade B. Meaning that 89.0% of screws were fully contained within the pedicle (GRS A), and 10% had a minor cortical breach <2 mm (GRS B). The median deviation from the planned trajectory was 0.2 mm (axial IQR: 0.1 to 0.5 mm; sagittal: IQR: 0.1 to 0.4 mm), in both axial and sagittal planes.
Conclusion:
This study showed that image registration of CT-like 3D MRI for robotic-assisted spine surgery is technically feasible and that accurate pedicle screw placement can be achieved without preoperative CT. Each CT-like 3D MRI was successfully registered for robotic navigation.
Clinical Significance:
The results suggest that CT-like 3D MRI has the potential to be a radiation-free alternative for preoperative planning and navigation in lumbar spine instrumentation procedures.
