Contour Map Point Distribution and Surgeon Experience Level Affect Accuracy of Surgical Navigation in a Pilot Study

Abstract

Background: Reliable use of surgical navigation depends upon the registration process. The gold standard is paired-point registration with bone-anchored fiducials, but contour-map registration is more practical. Surgeons may employ variable contour maps and less experienced team members often perform this critical step. The impact of these practices on target registration error (TRE) is not well-studied. Methods: A dry lab set-up consisting of a navigation system (Fusion ENT, Medtronic, Jacksonville, FL) and a sinus phantom with 2 mm radiopaque spheres in the sphenoid and ethmoid regions was developed. A CT (0.625 mm slice thickness) was obtained. Registration was performed with a contour-based protocol. Accuracy was determined using the software’s distance measurement tool. Registration was performed with narrow-field (NF; forehead points medial to the mid-pupillary line) and wide field (WF; entire forehead) contour maps. An experienced rhinologist and a resident surgeon performed each registration in triplicate and TRE at the sphenoid and ethmoid markers was measured in triplicate. Results: WF mapping had a lower TRE than NF (1.09 mm [95% Confidence Interval (CI) 0.96-1.22] vs 1.68 mm [95% CI 1.50-1.86]). The experienced surgeon had a lower TRE compared to the resident (1.21 mm [95% CI 1.08-1.34] vs 1.54 mm [95% CI 1.35-1.74]). Conclusions: In this navigation model, wide field mapping offers better accuracy than narrow-field mapping, and an experienced surgeon seemed to achieve better accuracy than a resident surgeon. These observations have potential implications for the use of this technology in the operating room.