Key Takeaways
- In the first same-patient, live-surgery comparison of both tracking modalities using surface registration, electromagnetic navigation required significantly less registration time than optic navigation (33.2 ± 2.4 s vs 58.9 ± 4.2 s; p < 0.05), while total surgical time did not differ between sides.
- Navigation error (NE) was clinically acceptable in both systems, with maximal mean deviations of 1.7 mm (optic) and 1.4 mm (electromagnetic) — both well within the generally accepted < 3 mm clinical threshold, and with no significant difference between systems at any measured landmark (p > 0.05).
- Accuracy was best in the medial-lateral (ML) axis (lamina papyracea 0.6 ± 0.4 mm optic; 0.6 ± 0.6 mm electromagnetic) and significantly worse in the anterior-posterior (AP) and cranial-caudal (CC) axes (p < 0.05, Bonferroni), implying higher risk near the skull base than near the orbit.
- Surface registration consumed < 2% of total operative time (1.58% optic; 0.98% electromagnetic), supporting its routine use as a fast, non-invasive registration strategy.
Clinical Context
Functional endoscopic sinus surgery (FESS) has largely replaced traditional open approaches (e.g., Caldwell-Luc) for chronic rhinosinusitis, but retains a definite risk of orbital and intracranial complications. Computer-aided surgery (CAS) — image-guided navigation — was introduced to help surgeons identify anatomic landmarks in real time and reduce catastrophic injury to the orbit, internal carotid artery, and skull base.
Two tracking modalities dominate clinical practice:
- Optic navigation triangulates infrared-tracked markers on instruments and the head reference frame. It is highly accurate but requires an unobstructed line of sight, so continuous guidance is frequently interrupted in a cluttered operating field.
- Electromagnetic navigation uses coil arrays that sense changes in a low-energy electromagnetic field, providing uninterrupted localization without line-of-sight constraints — advantageous in a crowded OR.
Registration — the correlation of preoperative CT/MRI to the patient's physical space — is the critical determinant of accuracy. Surface registration (matching unique facial contours) avoids the time-consuming placement of fiducial markers or landmark matching, at a modest cost in accuracy. Prior comparisons of optic vs electromagnetic systems used fiducial-marker registration; this study is the first to compare both modalities using surface registration in the same patients during live surgery, providing an internally controlled, three-dimensional accuracy analysis.
Clinical Evidence & Cohort Breakdown
Thirty patients (24 men, 6 women; age 20–62 years, mean 42) with bilateral chronic paranasal pansinusitis (Lund-Mackay total score > 15; Meltzer endoscopic grade III–IV polyposis) underwent bilateral FESS using a single dual-modality navigation platform (Medtronic S7). Each patient was randomized (random number table) so that one side used optic navigation and the contralateral side used electromagnetic navigation, with the same surgical team operating throughout.
| Parameter | Optic Navigation | Electromagnetic Navigation | p-value |
|---|---|---|---|
| Unilateral Lund-Mackay score | 9.7 ± 1.5 | 9.3 ± 1.5 | NS |
| Surface registration time (s) | 58.9 ± 4.2 | 33.2 ± 2.4 | < 0.05 |
| Surgical procedure time (min) | 62 ± 11.5 | 56.7 ± 11.1 | NS |
| Registration as % of operative time | 1.58% | 0.98% | — |
| NE — anterior ethmoid roof (AE), mm | 1.7 ± 1.0 | 1.3 ± 0.7 | NS |
| NE — posterior ethmoid roof (PE), mm | 1.5 ± 0.9 | 1.2 ± 0.8 | NS |
| NE — lamina papyracea (LP), mm | 0.6 ± 0.4 | 0.6 ± 0.6 | NS |
| NE — superior turbinate insertion (ST), mm | 0.6 ± 0.5 | 0.7 ± 0.5 | NS |
| NE — sphenoid ostium (OS), mm | 1.2 ± 0.7 | 1.1 ± 0.6 | NS |
| NE — upper border of choana (UC), mm | 1.4 ± 0.8 | 1.4 ± 0.6 | NS |
| Operative complications | 0 | 0 | — |
Axis-level findings: Within each system, NE differed significantly across the six landmarks (one-way ANOVA, p < 0.05). The ML-axis landmarks (LP, ST) were significantly more accurate than the AP- and CC-axis landmarks (AE, PE, OS, UC) in both systems (Bonferroni, p < 0.05). No significant differences were found among the AE, PE, OS, and UC landmarks, nor between LP and ST. Critically, no paired comparison between optic and electromagnetic systems reached significance at any landmark (independent t-test, p > 0.05).
Intraoperative IONM Protocol & Technical Pearls
The following step-by-step workflow reflects the navigation setup and measurement protocol used in this paired study and can be adapted to any dual-modality CAS platform.
- Preoperative imaging acquisition. Obtain a sinus-protocol CT at 1 mm slice thickness and transfer the dataset to the navigation unit before the case. Thin-slice imaging underpins the accuracy of any registration method.
- Head reference frame placement. After induction of general anesthesia, secure the head reference frame (with patient tracker) at the center of the forehead using an elastic strap. Confirm firm attachment and verify that its position relative to the head will remain stable throughout registration and surgery. Exclude patients with fragile or circulation-compromised skin at the mounting site.
- Localizer setup — optic system. Position the infrared camera approximately 1.75 m from the patient reference and confirm a clear, unobstructed line of sight to the optical markers on all tracked instruments. Anticipate that line-of-sight interruptions will degrade continuous guidance.
- Localizer setup — electromagnetic system. Mount the low-energy field emitter on a holder attached to the operating table frame, positioned at least 20 cm above the table and aimed directly at the patient's nose from a 20 cm distance. Ensure no part of the holder extends into the electromagnetic field, and connect all instrument tracker cables to the navigation unit ports.
- Surface registration. Distribute registration points across the nose, forehead, and orbital rim using the surface-matching algorithm (Synergy Cranial Software). Collect a sufficient point cloud for automatic 3D matching to the CT dataset, then confirm the instrument-to-image correlation. Record registration time for each modality.
- Maintain reference stability. Avoid moving or tilting the patient's head or the operating table during surgery. If bulb-press testing is performed to check lamina papyracea integrity, keep the head frame attached and stable until navigation is complete.
- Three-axis error measurement. After clearing abnormal sinonasal tissue to expose the skull base, lamina papyracea, and sphenoid sinus, touch each predetermined landmark with the navigated probe. Measure NE as the perpendicular distance from the navigated crosshair tip to an imaginary plane through the CT landmark:
- CC axis — sagittal plane, using the central points of the anterior and posterior ethmoid roofs (AE, PE).
- ML axis — coronal plane, using the most medial point of the lamina papyracea (LP) and the superior turbinate insertion (ST).
- AP axis — axial plane, using the sphenoid ostium (OS) and upper border of choana (UC).
- Repeat and average. Perform each distance measurement three times without re-registering, and record the mean as the NE for that landmark. Express all values as absolute distances for analysis.
- Interpret with axis-specific caution. Remember that a zero error in one axis does not exclude error in the others; treat AP and CC deviations as the dominant risk near the skull base.
Clinical Implications for Surgeons & IONM Specialists
- Both modalities are clinically equivalent in accuracy. With maximal mean errors of 1.7 mm (optic) and 1.4 mm (electromagnetic) — comfortably below the 3 mm clinical threshold — either system is acceptable for navigated FESS. Modern electromagnetic hardware has closed the historical accuracy gap with optic tracking, likely through improved resistance to ferromagnetic distortion.
- Electromagnetic navigation offers a workflow advantage. Surface registration was ~26 seconds faster with electromagnetic tracking, and continuous, line-of-sight-free localization is a meaningful ergonomic benefit in a cluttered OR. The authors recommend electromagnetic navigation for CAS on this basis.
- Registration time is negligible either way. At < 2% of total operative time, surface registration is a practical, non-invasive alternative to fiducial-marker or landmark registration, though it carries a small accuracy penalty from facial soft-tissue displacement and reference-frame drift.
- Respect the axis-dependent error profile. Accuracy is best in the ML direction (near the lamina papyracea) and worse in the AP and CC directions. Dissection near the skull base therefore carries greater navigational risk than dissection near the orbit — surgeons should slow down and cross-check anatomic landmarks when working in the AP/CC axes.
- Navigation is an adjunct, not a substitute. Thorough knowledge of complex sinonasal anatomy and surgical technique remains essential; navigation should confirm, not replace, anatomic judgment. Consider re-registration if intraoperative anatomical drift is suspected.
Reference
Chang CM, Jaw FS, Lo WC, Fang KM, Cheng PW. Three-dimensional analysis of the accuracy of optic and electromagnetic navigation systems using surface registration in live endoscopic sinus surgery. Rhinology. 2016;54(1):88-94. DOI: 10.4193/Rhino15.131
Medical Disclaimer: This article is provided for educational and informational purposes only and does not constitute medical advice. Clinical decisions regarding surgical technique, navigation system selection, and patient management should be made by qualified healthcare professionals based on individual patient circumstances and institutional protocols. The summarized findings reflect the cited study and should be interpreted alongside the broader peer-reviewed literature.