Purpose: Real-time surgical navigation has been increasingly applied in implant placement. The initial registration procedures were found to substantially affect the overall accuracy of the system for zygomatic implant placement, but the exact number and distribution of fiducial markers was yet undetermined. This study aimed to determine the minimal numbers and optimal distributions of fiducial markers to achieve clinically acceptable accuracy in surgical navigation, through systematically analyzing the effects of different setups of fiducial markers on target registration errors (TRE).
Materials and methods: A maxillary phantom with bone-anchored fiducial markers was scanned using cone beam computed tomography, followed by data processing on Brainlab, a commercially available navigation system. A total of 10 mini-screws were inserted in the edentulous maxilla for configuration of fiducial markers, with another two mini-screws as implant targets to assess TRE in zygomatic bone. Data were then collected in nine configurations with distinct fiducial numbers and positions. Statistical analyses were performed with SPSS.
Results: Accuracy of the surgical navigation system was found to depend on both the numbers and positions of fiducial markers. No significant difference was observed in accuracy among groups with eight fiducials and with polygon span distribution (P > .05). When the fiducial numbers decreased to less than six, the markers inserted in a regular triangle area were more precise than in an inverse triangle configuration. When the number of fiducials was five with a polygonal distribution, a low TRE value of 0.59 mm was detected, which was comparable to the accuracy with more than eight fiducials in the study.
Conclusion: A scattered distribution with a polygon span with at least five fiducial markers in the edentulous maxilla for registration seems to achieve an acceptable TRE value with a high accuracy for navigation in zygomatic implant placement.