Assessment of plan–execution fidelity in a fully digital workflow using custom-made plates for mandibular fracture repair: a prospective pilot study
摘要
Virtual surgical planning (VSP) and custom-made plates (patient-specific implants, PSIs) are increasingly used in mandibular fracture repair, but quantitative data on plan–execution fidelity remain limited. Plan–execution fidelity refers to the geometric agreement between the virtually planned and postoperative mandibular morphology after image registration. The aim of this pilot study was to quantify plan–execution fidelity and to describe patient-reported recovery in a fully digital, custom-plate workflow for non-comminuted mandibular fractures, while generating hypotheses about plate design.
MethodsSix patients with non-comminuted mandibular body or parasymphyseal fractures underwent open reduction and internal fixation (ORIF) using a fully digital workflow. Plan–execution fidelity (the primary feasibility outcome) was assessed as the mean absolute error (MAE) of 10 predefined measurements after iterative closest point (ICP) registration on the contralateral ramus and condyle (condylar metrics excluded a priori). Measurements were performed by three blinded observers. The Oral Health Impact Profile-14 (OHIP-14) was recorded preoperatively and at 1, 4, and 12 weeks. Analyses are purely descriptive; no inferential statistics are reported.
ResultsIn this six-patient cohort, the primary (patient-level) MAE ranged from 0.25 to 0.81 mm (mean 0.48 mm, SD 0.21 mm). At the level of individual measurements (60 readings from the same six patients), the MAE was 0.45 mm. OHIP-14 total scores improved from 48.2 (SD 5.6) preoperatively to 5.2 (SD 4.1) at 12 weeks. Plate design, indication, and chronological order were structurally confounded; therefore, no comparative interpretation is possible.
ConclusionsIn this pilot cohort, a fully digital workflow using custom-made plates was associated with submillimetre plan–execution deviation (mean 0.48 mm) within the registration framework used. The workflow was technically feasible and showed close correspondence between the virtual plan and postoperative morphology, although this geometric agreement should not be interpreted as independent anatomical accuracy. These preliminary findings provide a basis for future controlled studies aimed at validating clinical relevance, workflow efficiency, and cost-effectiveness in larger patient cohorts.
Trial registrationClinicalTrials.gov NCT07499726; registered 24 March 2026 (2026-03-24), retrospectively, after initiation of recruitment.