Evaluation of the Effectiveness of an AI-assisted, Dual-Template Workflow in Improving the Accuracy of Tooth Autotransplantation

Objective: To test whether an AI-assisted, dual-template workflow improves plan–delivery accuracy in tooth autotransplantation versus a replica-only freehand approach. Methods: Twenty-four consecutive cases were assigned to a guided workflow (n = 12) or freehand (n = 12). In the guided group, AI-driven CBCT segmentation informed a socket-preparation guide, a socket-margin guide, and a printed donor-tooth replica; the freehand group used the replica alone. Immediate postoperative CBCT was rigidly registered to the plan. The prespecified primary endpoint was buccolingual angular deviation (BLAD). Secondary, exploratory endpoints included centre deviation distance (CDD), index-point distances at crown and root (IPDD-C/IPDD-R), long-axis (LAD), mesiodistal (MDAD) and rotational (RAD) deviations, and Dice similarity coefficient (DSC). Group differences were tested with t-tests or Mann–Whitney U tests. Results: The guided workflow yielded significantly lower BLAD than the freehand approach. Secondary analyses showed concordant reductions in rotational deviation and linear centring errors, with greater volumetric plan–achievement overlap. Long-axis deviation showed a modest, non-significant improvement, and mesiodistal deviation did not differ between groups. Conclusions: An AI-assisted dual-template workflow improves plan–delivery fidelity in tooth autotransplantation, with the largest gains in buccolingual orientation and rotational control, alongside better coronal centring and increased volumetric concordance. Apical deviation was also reduced, whereas mesiodistal alignment was unchanged. Clinical significance: By constraining trajectory and socket margins, the guided workflow standardizes preparation and supports more predictable positioning of the transplanted tooth.