A comparative in vitro study on the accuracy of robotic arm mode versus dynamic navigation mode of an autonomous dental implant robot in the posterior mandibular region

Abstract

To investigate the application efficacy of the robotic arm mode versus the navigation mode of an autonomous dental implant robot in posterior mandibular implant surgery. A total of 80 custom-designed resin models were stratified and equally divided into an experimental group (robotic arm mode, n = 40) and a control group (dynamic navigation mode, n = 40). Implants were placed using the respective robotic operating modes. Postoperative cone-beam computed tomography (CBCT) scans were performed to measure the spatial deviations between the actual implant positions and the preoperatively planned positions, comparing accuracy across different unilateral posterior mandibular defect classifications. Data from the 120 placed implants were aggregated to the model level ( N = 80). In the robotic arm mode group, the entry point deviation, endpoint deviation, and angular deviation were 0.68 ± 0.18 mm, 0.75 ± 0.24 mm, and 1.26 ± 0.44°, respectively. In the dynamic navigation mode group, these deviations were 1.01 ± 0.27 mm, 1.12 ± 0.35 mm, and 2.34 ± 0.54°, respectively. (1) Under standardized in vitro conditions, both the robotic arm mode and the dynamic navigation mode achieved clinically acceptable implant accuracy in simulated posterior mandibular defects. (2) However, the accuracy of robotic arm mode-assisted implant placement was significantly superior to that of dynamic navigation mode, suggesting that the robotic arm mode reflects the anticipated precision advantages of automated mechanical execution over manually controlled navigation in a controlled laboratory setting.