Influence of different registration modalities on navigation accuracy in ear, nose, and throat surgery depending on the surgical field

Objectives/Hypothesis: Various invasive and noninvasive registration methods have been established in the past for intraoperative navigation. The present study compared the registration and navigation accuracy of three different registration modalities in anatomical locations of special interest for ear, nose, and throat surgery. Study Design: Prospective experimental phantom study. Methods: Four skull models were individually fabricated with a three‐dimensional printer based on the patient's computed tomography data sets and fitted with an individual customized silicone skin. Three different registration modalities were examined: 1) invasive marker (IM), 2) oral splint (OS), and 3) laser scan (L). Accuracy measurements were assessed by targeting 26 titanium screws placed over the skull. The overall accuracy and the target registration error for eight selected anatomical locations were measured. Results: Mean accuracy was 0.67 ± 0.1 mm (quadratic mean ± standard deviation) for IM, 0.98 ± 0.16 mm for OS, and 1.3 ± 0.12 mm for L. The greatest differences in accuracy were found on the mastoid with best accuracy for IM (0.59 ± 0.2 mm; P < .05 vs. OS and L), followed by OS (1.23 ± 0.41 mm; P < .05 vs. L), and L (1.88 ± 0.45 mm). In contrast, only small differences in accuracy were detected in the anterior skull base between the registration modalities (IM 0.75 ± 0.21 mm, OS 0.71 ± 0.27 mm, L 0.93 ± 0.34 mm). Conclusions: L and OS meet accuracy requirements in the midface and anterior skull base. OS is superior to L with navigation accuracies comparable to marker registration. However, neither method meets the high precision requirements for lateral skull base surgery. Laryngoscope, 2010