SlicerOrbitSurgerySim: Orbital Plate Fit Analysis

• SlicerOrbitSurgerySim is an open-source extension for 3D Slicer that supports interactive virtual registration and quantitative analysis of preformed orbital plates.
• It integrates advanced imaging, landmark-based alignment, and mesh-driven distance computation to provide objective, reproducible surgical metrics and 3D visualizations.
• The platform facilitates comparative studies of implant designs across vendors and patient anatomies, enhancing preoperative planning and research in orbital reconstruction.

SlicerOrbitSurgerySim is an open-source extension for 3D Slicer designed to enable interactive virtual registration, quantitative fit analysis, and objective comparison of preformed orbital plates in the context of orbital surgery planning. The platform integrates medical imaging data, user-guided and metric-driven implant positioning algorithms, and reproducible mesh-based distance computation and visualization. By supporting reproducible quantitative plate-to-orbit metrics and comparative studies across plate vendors, sizes, and patient anatomies, SlicerOrbitSurgerySim provides a standardized and transparent foundation for surgical planning, research, and education in orbital reconstruction and implant design (Zhang et al., 22 Dec 2025).

1. System Architecture and Modules

SlicerOrbitSurgerySim is implemented as a modular extension within 3D Slicer (5.10.0+), structured as three primary modules and one optional utility:

• MirrorOrbitRecon: Automates orbital mirroring (contralateral to defect), performs rigid and optional Coherent Point Drift (CPD) deformable alignment, and outputs reconstructed orbit meshes.
• PlateRegistration: Enables sequential, surgeon-driven alignment of preformed plate models to reconstructed orbital anatomy via landmark-based rigid registration, posterior-stop anchoring, interactive refinement, and real-time collision checking.
• FitMetrics: Computes per-vertex and curve-based plate-to-orbit distances, generates quantitative fit metrics, produces 3D heatmaps and 2D overlays, and exports pointwise and summary statistics.
• DentalSegmentator (optional): Supports automatic segmentation of skull and orbit structures from CT.

Key dependencies include SlicerMorph (CPD, I/O), ModelToModelDistance (VTK-based pointwise distances), SimpleITK or Eigen (ICP/rigid alignment), and the VTK collision detection filter. Data inputs consist of DICOM CT (imaging), STL/OBJ/PLY (plate/mesh), with outputs in CSV (metrics), STL/VTP (labeled meshes), and PNG (screenshots). The internal dataflow is orchestrated via the Medical Reality Markup Language (MRML) node system for scene management and provenance (Zhang et al., 22 Dec 2025).

2. Virtual Registration Workflow

Plate registration comprises a structured three-stage process:

1. Landmark-Based Rigid Alignment The user places 3–5 corresponding landmarks on the plate and reconstructed orbit. The optimal rotation $R$ and translation $t$ minimizing

$$\min_{R \in SO(3),\,t \in \mathbb{R}^3} \sum_{i=1}^N \| R\, \mathrm{plate\_pts}_i + t - \mathrm{orbit\_pts}_i \|^2$$

is solved by Horn’s method or Slicer's LandmarkTransform.

1. Posterior-Stop Anchoring Single-point correspondences ($p_\mathrm{stop}$ on the plate, $q_\mathrm{stop}$ on the orbit) define anchoring. An offset $\Delta t$ ensures $$R\,p_\mathrm{stop} + (t + \Delta t) = q_\mathrm{stop}$$. Post-anchoring, subsequent manual manipulations rotate about $q_\mathrm{stop}$ (Zhang et al., 22 Dec 2025).
2. Interactive Refinement and Collision Checking Surgeons refine placement with Slicer’s 3D transforms widget and receive real-time feedback via VTK collision detection: intersecting vertices are labeled and percentages displayed. Registration is iterated until both visual and collision-based criteria are met. Optional ICP refinements are supported for rigid-only convergence; CPD is reserved for mirror-based reconstructions, not plate registration (Zhang et al., 22 Dec 2025).

Pseudocode for the registration process is provided in the original documentation.

3. Quantitative Plate-to-Orbit Metrics

SlicerOrbitSurgerySim computes both vertex-wise and edge-specific distance metrics between transformed plate and orbit meshes:

• Point-to-Surface Distances: For each plate vertex $p_i$, the minimal Euclidean distance to the orbit mesh $O$ is

$d(p_i, O) = \min_{q \in O} \| p_i - q \|$

• Statistical Summaries: Mean $(\bar{d})$, maximum $(d_\mathrm{max})$, root-mean-square $(d_\mathrm{rms})$, standard deviation $(\sigma)$, and above-threshold area $(A_{>T})$ (based on 1-ring vertex area) summarize overall fit. Distance histograms are binned over user-specified intervals.
• Edge-Specific Metrics: Manually drawn edge curves are projected onto the orbit. For each sampled curve point $c_j$, projection $q_j$ minimizes $\|c_j - q\|$ over the orbit, computed for all $j$ on each of five relevant anatomical edges.

Thresholds of 1 mm (excellent), 2 mm (acceptable), and greater (poor) are available for fit quality assessments, motivated by pilot clinical analysis (Zhang et al., 22 Dec 2025).

4. Visualization and Analytical Tools

Visualization tools are tailored for both qualitative and quantitative assessment:

• 3D Heatmaps: The plate’s vertices are colored by signed distance (colormap: blue $<-$5 mm, green 0 mm, red $+5$ mm).
• 2D Slice Overlays: Markup curves and their projections are rendered as overlays on cross-sectional anatomy, facilitating scrutiny of implant-to-bone proximity in all planes.
• Cross-sectional and Statistical Plots: Histograms and scatterplots of distances can be visualized with Slicer's Plot view or exported for external processing.
• Module User Interfaces: Each module is a self-contained Slicer sidebar tab with controls for parameter tuning, curve-drawing, collision reporting, and exporting of metrics and heatmap meshes.

5. Population-Level Analysis and Export

SlicerOrbitSurgerySim supports population-level comparative studies and multi-plate evaluation:

• Per-patient, per-plate CSV output includes all per-vertex and summary statistics.
• High-throughput analysis is performed by concatenating exported CSVs in R or Python/pandas for visualization (box plots, violin plots) and inference (ANOVA, post-hoc Tukey HSD).
• Edge-specific and platewide metrics allow fine-grained analyses across vendors, sizes, and patient characteristics.
• Sample data and tutorials facilitate reproducible research design and cohort-level analytics. For more complex designs, export to external statistical environments is recommended (Zhang et al., 22 Dec 2025).

6. Workflow and Practical Use

The canonical workflow for SlicerOrbitSurgerySim is:

1. Install the extension from the 3D Slicer Extension Manager.
2. Load DICOM CT imaging; (optionally) apply DentalSegmentator or other segmentation to obtain orbit/skull meshes.
3. Use MirrorOrbitRecon for mirroring/reconstruction; approve or refine results.
4. Load plate meshes (STL/OBJ/PLY).
5. Use PlateRegistration for sequential alignment (landmark, anchoring, refinement, collision evaluation).
6. In FitMetrics, compute and export quantitative distance metrics, generate 3D colored overlays, and edge-specific analyses.
7. Repeat for all plates under consideration; aggregate outputs for comparison.
8. For population-level analysis, concatenate exported CSVs and apply relevant statistical methods.

Preprocessing recommendations include cropping to orbit region, applying smoothing filters, and mesh decimation to optimize interactivity (Zhang et al., 22 Dec 2025).

7. Validation, Impact, and Available Resources

A pilot validation cohort (14 patients, 4 plate designs) demonstrated use-case feasibility and discriminatory power for clinical and research questions:

• Mean plate-to-orbit distances were in the range 1.4–1.7 mm, with best-fit values approaching 1.1 mm.
• Edge-specific metrics identified design-dependent fit differences, such as improved adaptation of Vendor B's convex posterior floors to certain anatomies.
• All datasets, output files, and guided tutorials (including annotated exemplar cases) are freely available in the Slicer Sample Data repository under “OrbitPlateFit.”
• Provided self-guided tutorials walk users through the complete process from segmentation through comparative analysis.

The system is designed to facilitate improved preoperative decision-making, objective evaluation of implant design and placement, reproducible cross-institution studies, and collaborative education (Zhang et al., 22 Dec 2025).