Virtual Occlusion Accuracy in Orthognathic Surgery: Challenges and Critical Limitations in Segmental Maxillary Planning.

BACKGROUND

Dental occlusion is critical in orthognathic planning. Although manual articulation of casts was considered the reference standard, digital workflows are adopted.

PURPOSE: To assess linear and angular discrepancies between virtual occlusion (VO) and manual reference occlusion (MrO) to determine VO accuracy.

STUDY DESIGN, SETTING, AND SAMPLE: A prospective cohort study included subjects undergoing orthognathic surgery at the Division of Maxillo-Facial Surgery, University of Turin, between January 2024 and December 2025. Patients undergoing distraction osteogenesis, lacking sufficient antagonistic dentition or imaging, or presenting atypical segmentation patterns were excluded.

PREDICTOR VARIABLES: Occlusal planning modality (virtual, manual); surgery type (nonsegmental, two-piece, and three-piece).

MAIN OUTCOME VARIABLES: VO accuracy. In nonsegmental surgery, for each linear and angular parameter, VO was considered accurate when the mean absolute discrepancy from MrO did not statistically significantly exceed the manual interoperator discrepancy (MrO vs an independent manual occlusion). In segmental surgery, VO was considered accurate when linear measurements showed no statistically significant difference from MrO and mean absolute angular discrepancies were <2 °. Discrepancies were measured using iterative closest point registration.

COVARIATES: Demographic (age, sex).

ANALYSES: Paired t-tests (α = 0.05) compared: (1) VO with manual interoperator discrepancies; (2) intra-arch measurements between MrO and VO.

RESULTS

Eighty four patients were screened; 57 (67.9%) subjects were included. The mean age was 26.0 (5.6) years; 28 (49.1%) were male. The sample included 39 (68.4%) nonsegmental, 10 (17.5%) two-piece, and 8 (14.0%) three-piece cases. In nonsegmental surgery, VO discrepancies statistically significantly exceeded manual inter-operator discrepancies for cranio-caudal translation (P < .001), pitch (P = .002), roll (P < .001), and yaw (P = .007). VO was accurate for anteroposterior (P = .09) and lateral (P = .10) translations. In segmental surgery, VO was accurate for linear measurements (two-piece: P = .7; P > .9; three-piece: P = .4; P = .6). In two-piece surgery, VO was accurate for pitch, 1.66 ° (1.28 °), but not for yaw, 2.44 ° (1.47 °) or roll, 2.42 ° (1.51 °). In three-piece surgery, VO was accurate for lateral segments pitch, 1.51 ° (0.64 °); other angular discrepancies exceeded 2 ° (lateral segments: yaw 3.07 ° (0.38 °), roll 2.86 ° (0.27 °); incisal segment: yaw 2.22 ° (0.13 °), roll 2.52 ° (0.13 °), pitch 2.58 ° (0.22 °)).

CONCLUSIONS AND RELEVANCE: VO accuracy varied across linear and angular parameters; limitations primarily affected angular parameters.