Trueness analysis of a robot-assisted system for maxillary supernumerary anterior tooth removal: an in vitro study

The present study explored the trueness of active robotic system-assisted cortical-window osteotomy and tooth sectioning relative to conventional freehand surgery, employing a labial approach to remove impacted maxillary midline supernumerary teeth. Standardized three-dimensional maxillary replicas containing an impacted midline supernumerary tooth were randomized into an Active Robot-assisted Group (ARG) or a Freehand Group (FHG). Preoperative CBCT and intraoral scans were merged to plan osteotomy and tooth sectioning trajectories. Parameters analyzed included cortical-window formation and tooth segmentation trueness, bone injury, and operative duration. The ARG achieved significantly higher Dice similarity coefficients (0.93 vs. 0.71, p < 0.001) and lower centroid deviations (0.31 mm vs. 1.15 mm, p < 0.001) than the FHG. Both excessive and insufficient bone removal were markedly reduced in the ARG group (p < 0.001). The root mean square (RMS) deviation (0.19 mm vs. 0.96 mm, p < 0.001) and angular deviation (5.70° vs. 17.90°, p < 0.001) were significantly smaller in the ARG group, while the depth of bone injury was lower (0.00 [0.00-0.77] vs. 0.87 [0.45-1.51], p = 0.039), and the operative duration was slightly longer (160 s vs. 132 s). Robot-assisted extraction of impacted maxillary midline supernumerary teeth demonstrated greater trueness and reduced collateral bone injury than conventional freehand surgery. The robotic system enhanced surgical safety, reproducibility, and trueness. These in vitro findings demonstrate the technical feasibility of active robotic systems for minimally invasive, high-trueness extraction of supernumerary teeth.