Skip to content
ENT SurgicalTech & Protocols

Registration-Free Navigation in Orbital and Craniofacial Surgery: Accuracy of Splint-Borne Dynamic Reference Frames

Cadaver study comparing registration-free splint-borne DRF navigation accuracy across electromagnetic and optical tracking versus fiducial and surface-based registration.

IP

Editorial Team

Clinical evidence review

5 min read
Source:Scientific Reports (2021)DOI10.1038/s41598-021-97706-5

Key Takeaways

  • A registration-free workflow using a pre-planned, splint-borne Dynamic Reference Frame (DRF) eliminated the intraoperative registration step entirely, achieving a Target Registration Error (TRE) intercept of 0.84 mm with electromagnetic (EM) tracking—the most accurate of all tested approaches at the infraorbital rim.
  • The same registration-free workflow performed poorly with optical tracking, with an intercept of 1.16 mm and 41% of measurements exceeding 2 mm TRE, versus only 11% for EM registration-free navigation.
  • Bone-anchored fiducial registration remained accurate at the infraorbital rim (0.97 mm optical; 0.91 mm EM) but degraded more steeply with increasing target distance than soft-tissue registration (slope 0.0049 vs 0.0013 mm per mm).
  • The unexplained accuracy divergence between tracking techniques (P < 0.001 for the EM × registration-free interaction) currently limits clinical implementation of this otherwise time-efficient, non-invasive, radiation-free method.

Clinical Context

Accurate intraoperative registration is the cornerstone of reliable surgical navigation in orbital and craniofacial procedures. The ideal registration method would be non-invasive, rapid, logistically simple, automatic (and therefore user-independent), universally applicable, compatible with both optical and electromagnetic tracking, free of additional radiation exposure, and—above all—accurate.

Conventional approaches each carry trade-offs. Bone-anchored fiducial markers and splint-based point registration introduce invasiveness and user dependency. Surface-based registration via touch or laser scanning adds accuracy variability and workflow burden. Automatic image registration eliminates user dependency but requires intraoperative imaging, introducing radiation exposure and extended operative time.

This study, conducted at high-volume tertiary surgical centers (Amsterdam UMC and Radboud University Medical Center), proposes a registration-free dental splint-based method that removes both user dependency and the need for additional intraoperative imaging. The maxillary dental model is fused to preoperative imaging using commercially available virtual planning software, and a virtual DRF is designed directly on the patient's fused maxillary dentition. During surgery, the splint containing the reference frame is seated on the dentition, and because the DRF pose is known by design, no registration procedure is required.

Clinical Evidence & Cohort Breakdown

The accuracy of the registration-free workflow was evaluated in a cadaver setup (five dentulous cadaver heads) and compared against bone-anchored fiducial registration (optical and EM) and surface-based registration (EM). A total of 3,496 data points were analyzed, including 1,396 registration-free measurements (4 missing, 0.3%). Fourteen PEEK Allen screws served as ground-truth targets, and TRE was modeled with a linear mixed model.

Registration Method Tracking Technique TRE Intercept (mm) Slope (mm TRE per mm distance) Measurements > 2 mm TRE
Bone-anchored fiducials Optical 0.97 0.0049 27%
Registration-free Optical 1.16 0.0039 41%
Bone-anchored fiducials Electromagnetic 0.91 0.0042 15%
Surface-based Electromagnetic 1.10 0.0013 17%
Registration-free Electromagnetic 0.84 0.0036 11%

The linear mixed model identified significant fixed effects for tracking technique, registration method, and target distance, including a significant EM × registration-free interaction (estimate −0.253, t = −8.87, P < 0.001). The registration-free EM approach outperformed all other registration approaches, whereas the registration-free optical approach was outperformed by bone-anchored fiducial optical registration and all EM registration approaches.

Intraoperative IONM Protocol & Technical Pearls

  1. Acquire preoperative imaging and an intraoral scan. Obtain a CT scan and a high-resolution intraoral scan (stereolithographic model) of the maxillary dentition. Fuse the dental model to the maxillary dentition of the CT scan in virtual planning software (e.g., IPS CaseDesigner).
  2. Design the splint with a 0.1 mm offset. Import the fused dental model into 3D modeling software (e.g., Blender) and design a splint fitting the dentition using a 0.1 mm offset to ensure proper fit and minimize positional deviation.
  3. Integrate the DRF geometry into the splint design. Incorporate either reflective markers (optical, resembling the Skull Reference Array) or an EM reference holder into the splint. Design extensions with rounded edges to maximize bending resistance and stiffness, since material flexibility can distort the true DRF position.
  4. Manufacture via 3D printing in rigid transparent material. Export designs in stl format and print with a PolyJet printer (e.g., VeroClear). Verify geometry and marker-sphere distances before use; all splint-borne DRFs must be recognized by the navigation system.
  5. Establish the DRF pose in image space preoperatively. Calculate the transformation from the splint to the navigation image space, accounting for differences in reference frame construction (voxel vs image space) and image orientation (RAS vs LPS) between planning and navigation software.
  6. Seat and secure the splint intraoperatively. Position the splint-borne DRF on the patient's dentition and secure it (e.g., with power chains). Repositioning of the splint is the primary act determining measurement outcome—verify stable, reproducible seating, particularly at the molar region.
  7. Prefer electromagnetic tracking for this workflow. Given the significant accuracy advantage of EM over optical tracking in the registration-free setting, select EM tracking when implementing this protocol until the source of the optical discrepancy is resolved.
  8. Account for target distance. Expect TRE to increase with distance from the splint centroid; the slope values above allow estimation of accuracy at remote orbital and craniofacial targets.

Clinical Implications for Surgeons & IONM Specialists

This registration-free workflow offers a genuinely time-efficient, non-invasive, radiation-free, and automatic alternative to conventional registration, with the potential to reduce user dependency and streamline operating room logistics. However, the striking and unexplained accuracy divergence between tracking techniques—EM registration-free navigation was the most accurate approach tested, while optical registration-free navigation was the least accurate—remains a critical barrier to clinical adoption.

Proposed contributors to the optical discrepancy include reduced stiffness of the extended reflective-marker arms, fit differences between the incisor and molar regions, and possible design or fabrication inaccuracies in the optical DRF. None fully accounts for the magnitude of the difference. Until this is resolved, teams implementing registration-free navigation should favor EM tracking and validate splint fit and DRF geometry rigorously. Bone-anchored fiducial registration remains a reliable benchmark at the infraorbital rim, though its accuracy degrades more steeply with increasing target distance than soft-tissue registration.

Reference

Schreurs R, Baan F, Klop C, Dubois L, Beenen LFM, Habets PEMH, Becking AG, Maal TJJ. Registration-free workflow for electromagnetic and optical navigation in orbital and craniofacial surgery. Scientific Reports. 2021;11:18080. doi:10.1038/s41598-021-97706-5

Medical Disclaimer: This article is provided for informational and educational purposes only and does not constitute medical advice. Clinical decisions regarding surgical navigation, registration techniques, and intraoperative monitoring should be made by qualified healthcare professionals based on individual patient circumstances and institutional protocols.

Medical disclaimer

This protocol is provided for clinical education and reference only. It does not constitute medical advice, a standard of care, or a substitute for surgeon judgment, institutional policy, device labeling, or patient-specific assessment. Verify all stimulation parameters, electrode compatibility and indications against current manufacturer instructions before use. Read the full disclaimer.

Continue the protocol

Related guidance

View the ENT Navigation library