A hybrid robotic system for zygomatic implant placement based on mixed reality navigation
Xingqi Fan,Yuan Feng,Baoxin Tao,Yihan Shen,Yiqun Wu,Xiaojun Chen
DOI: https://doi.org/10.1016/j.cmpb.2024.108156
Abstract:Backgrounds: Zygomatic implant (ZI) placement surgery is a viable surgical option for patients with severe maxillary atrophy and insufficient residual maxillary bone. Still, it is difficult and risky due to the long path of ZI placement and the narrow field of vision. Dynamic navigation is a superior solution, but it presents challenges such as requiring operators to have advanced skills and experience. Moreover, the precision and stability of manual implantation remain inadequate. These issues are anticipated to be addressed by implementing robot-assisted surgery and achieved by introducing a mixed reality (MR) navigation-guided hybrid robotic system for ZI placement surgery. Methods: This study utilized a hybrid robotic system to perform the ZI placement surgery. Our first step was to reconstruct a virtual 3D model from preoperative cone-beam CT (CBCT) images. We proposed a series of algorithms based on coordinate transformation, which includes image-phantom registration, HoloLens-tracker registration, drill-phantom calibration, and robot-implant calibration, to unify all objects within the same coordinate system. These algorithms enable real-time tracking of the surgical drill's position and orientation relative to the patient phantom. Subsequently, the surgical drill is directed to the entry position, and the planned implantation paths are superimposed on the patient phantom using HoloLens 2 for visualization. Finally, the hybrid robot system performs the processed of drilling, expansion, and placement of ZIs under the guidance of the MR navigation system. Results: Phantom experiments of ZI placement were conducted using 10 patient phantoms, with a total of 40 ZIs inserted. Out of these, 20 were manually implanted, and the remaining 20 were robotically implanted. Comparisons between the actual implanted ZI paths and the preoperatively planned ZI paths showed that our MR navigation-guided hybrid robotic system achieved a coronal deviation of 0.887 ± 0.213 mm, an apical deviation of 1.201 ± 0.318 mm, and an angular deviation of 3.468 ± 0.339° This demonstrates significantly better accuracy and stability than manual implantation. Conclusion: Our proposed hybrid robotic system enables automated ZI placement surgery guided by MR navigation, achieving greater accuracy and stability compared to manual operations in phantom experiments. Furthermore, this system is expected to apply to animal and cadaveric experiments, to get a good ready for clinical studies.