Accuracy comparison between active robot and freehand techniques in modified mid-segment-first extraction of impacted mandibular third molars: an in vitro study.

This study compared the precision of cortical-window osteotomy and tooth sectioning performed with an active robotic system versus conventional freehand techniques during the modified mid-segment-first (MMSF) extraction of impacted mandibular third molars (M3Ms). The feasibility of applying active robotic technology to alveolar surgery was also assessed. Standardized mandibular replicas containing fully impacted horizontal M3Ms (n = 18 per group) were allocated into two groups: an active robot-assisted group (ARG) and a freehand group (FHG). For all specimens, preoperative CBCT scans were merged with intraoral digital impressions to design the osteotomy and tooth sectioning pathways. Primary endpoints were cortical bony window accuracy and tooth segmentation accuracy. Dice similarity scores were significantly higher in the ARG compared with FHG (0.90 ± 0.03 vs. 0.29 ± 0.09, p < 0.001). Root mean square (RMS) deviation (0.14 [0.10-0.22] mm vs. 1.39 [1.07-2.09] mm, p < 0.001) and angular deviation (2.02 [1.31-3.16]° vs. 19.48 [13.30-29.07]°, p < 0.001) were lower in the ARG group, as was bone injury depth (0.00 [0.00-1.15] mm vs. 2.21 [0.89-3.56] mm, p < 0.001). Both excessive and insufficient removal were significantly less frequent in robotic procedures (p < 0.001). Robotic-assisted MMSF extraction demonstrated greater precision, reduced invasiveness, and improved safety compared to freehand techniques in vitro. These findings support the preliminary feasibility of active robotic systems in alveolar surgery; however, as this was an in vitro study, further evaluation is needed to balance the gains in accuracy against the inherent limitations. This study provides the first quantitative in vitro evidence validating the technical feasibility of an active robotic platform for dentoalveolar surgery, laying the groundwork for future clinical applications.