Joachim Polligkeit

This preclinical investigation examined whether optical see-through augmented reality (AR) can transfer planned fronto-orbital cutting paths during simulated frontal debossing compared with patient-specific CAD/CAM cutting guides (CGs). Twenty printed skull replicas derived from one CT dataset were divided equally between workflows. Four superficial frontal or supraorbital grooves were created per skull and evaluated using postoperative CBCT and bidirectional closest-point distances. Overall deviation averaged the four trajectory means within each skull, giving 10 independent observations per group. Skull-level median [Q1, Q3] deviation was 0.543 [0.496, 0.588] mm with CGs and 1.098 [0.907, 1.333] mm with AR. Exact two-sided Mann–Whitney comparisons with Holm correction across overall accuracy, execution time, and four regions favored CGs for overall accuracy (adjusted p = 0.0208), superior frontal and right supraorbital accuracy, and execution time (adjusted p = 0.0029). The median [Q1, Q3] cumulative groove-execution time was 65.5 [61.5, 78.5] s with CGs and 152.0 [131.5, 208.8] s with AR. Mechanical guides deviated less with shorter groove-execution times in this standardized model, while operator and learning effects were not adjusted for. These findings concern surface trajectory reproduction in rigid models; clinical validation and assessment of the complete workflow are needed.

The accurate reconstruction of craniofacial defects requires the precise segmentation and mirroring of healthy anatomy. Conventional workflows rely on manual interaction, making them time-consuming and subject to operator variability. This study developed and validated a fully automated digital pipeline that integrates deep learning–based segmentation with algorithmic mirroring for craniofacial reconstruction. A total of 388 cranial CT scans were used to train a three-dimensional nnU-Net model for skull and mandible segmentation. A Principal Component Analysis–Iterative Closest Point (PCA–ICP) algorithm was then applied to compute the sagittal symmetry plane and perform mirroring. Automated results were compared with expert-generated segmentations and manually defined symmetry planes using Dice Similarity Coefficient (DSC), Mean Surface Distance (MSD), Hausdorff Distance (HD), and angular deviation. The nnU-Net achieved high segmentation accuracy for both the mandible (mean DSC 0.956) and the skull (mean DSC 0.965). Mirroring results showed minimal angular deviation from expert reference planes (mandible: 1.32° ± 0.71° in defect cases, 1.58° ± 1.12° in intact cases; skull: 1.75° ± 0.84° in defect cases, 1.15° ± 0.81° in intact cases). The presence of defects did not significantly affect accuracy. This automated workflow demonstrated robust performance and clinical applicability, offering standardized, reproducible, and time-efficient planning for craniofacial reconstruction.