Randomized controlled trial comparing AI-assisted digital and conventional orthodontics: Superior PAR reduction and occlusal outcomes

Background: Advances in digital orthodontics and artificial intelligence (AI) planning have the potential to enhance treatment precision, but randomized evidence based on the Peer Assessment Rating (PAR) index remains limited. Methods: In this single-center, parallel-group randomized controlled trial registered retrospectively in the Chinese Clinical Trial Registry (ChiCTR2500108499), 140 patients aged 12–35 years with Angle Class I malocclusion were randomized to receive an AI-assisted digital workflow (Digital and AI group) or conventional fixed appliances (Conventional group). PAR scores were assessed at baseline (T0), 6-month intervals (T1), and immediately after treatment completion (T2) by calibrated, blinded examiners following British Standards Institute criteria. Analyses followed the intention-to-treat principle, applying independent t-tests, χ²/Fisher’s exact tests, repeated-measures mixed-effects models, and multivariable linear regression. Effect sizes were expressed as mean difference (MD) or relative risk (RR) with 95% confidence intervals (CI). Results: Baseline PAR scores did not differ significantly between groups (MD = 0.63, 95% CI: –0.13 to 1.40; p = 0.105). At T2, the Digital and AI group had lower mean PAR scores (4.88 ± 0.45) than the Conventional group (7.81 ± 0.70; MD = 2.93, 95% CI: 2.73–3.13; p