gms | German Medical Science

Objective: Most cranial stereotactic- and spinal navigation- or robotic procedures use computed tomography (CT) for referencing. However, young patients could profit from reduced radiation. Therefore, we established a cranial robot-guided stereotactic biopsy and implantation protocol, based on MR-scans, only. For robot-guided pedicle screw implantation 3D-fluoroscopy was established for referencing. We will provide accuracy studies for the cranial- and spinal robotic surgery protocol.

Methods: In the cranial part of the study, robot-guided implantation of sEEG electrodes by the cranial robotic surgery assistant (ROSA-brain, MedTech/Zimmer-Biomet, Montpellier, France) were analyzed. Planning was performed on preoperative MRIs with the ROSA-software. Patients heads were fixed in a sharp clamp, referencing was performed by laser-scan of the face. During surgery the robotic arm steered the probe. After surgery the final position of the electrodes was identified on a postoperative CT scan, which was merged with the preoperative planning MRI scan. The accuracy was determined by calculating the target point error (TPE) and the entry point error (EPE) applying the Euclidian distance. In the spinal part of the study, three fresh-frozen cadaver were used. Preoperatively 3D-fluoroscopy (Ziehm Vision RD Vario 3D, Nuernberg, Germany) was performed and referenced to a tracker, fixed on the processus spinosus. Screw trajectories were planned with the ROSA-software. A total of 12 pedicle screws (Viper, Depuy Synthes, Zuchwil, Switzerland) were percutaneously implanted through a probe, steered by spine-robot (ROSA-spine, MedTech/Zimmer-Biomet). The final position of the pedicle screws was identified with 3D-fluoroscopy and compared to the preoperatively planned trajectories with the ROSA-software.

Results: A total of 40 sEEG electrodes were implanted in five patients. The mean TPE was 2.96 mm, the mean EPE 2.53 mm. We observed a difference in accuracy depending on the MRI field strength. The mean TPE (1.72 mm) in one patient who received a 1.5 Tesla MRI was better than the average TPE we achieved in our patients. No complications related were observed. In the spinal study, we observed accurate navigation and facilitated real time virtual visualization of the surgical procedure. The depth and position of the screw could be monitored by the surgeon during the placement procedure. All screws were placed in a correct transpedicular position, no major medial or lateral deviations or breaches of the pedicle wall were observed.

Conclusions: The cranial MR-based robotic protocol, as well as the spinal 3D-fluoroscopy-based robotic protocol, have shown equal accuracy compared with studies using CT for referencing. Both protocols are suitable to reduce radiation exposure of patients.