Related Experiment Video
Updated: Sep 25, 2025
Surgical Bone Implantation Technique for Rat Tibia Models of Diabetes and Osteoporosis
Published on: July 5, 2024
Influence of bone density on the accuracy of artificial intelligence-guided implant surgery: An in vitro study
Zhicong Chen1, Yun Liu2, Xin Xie3
1Graduate student, Department of Oral Implantology, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-Sen University, Guangzhou, Guangdong, PR China.
Statement Of Problem:
Artificial intelligence (AI) has been found to be applicable in medical tests and diagnostics. However, studies on the application of AI technology in oral implantology are lacking. In addition, whether bone density affects the accuracy of guided implant surgery has not been determined.
Purpose:
The purpose of this in vitro study was to determine the clinical reliability of an AI-assisted implant planning software program with an in vitro model. An additional goal was to determine the effect of bone density on the accuracy of static computer-assisted implant surgery (CAIS).
Material And Methods:
Ten participants with missing mandibular left first molars were selected for analysis, and surgical fully guided templates were designed by using an AI implant planning software program. Jaw models were produced in 3 filling rate groups (group L: 25%; group M: 40%; group H: 55%, higher filling rate with representatives of the denser simulated bone density) by 3-dimensional (3D) printing. The preoperative and postoperative positions of the implants were compared by measuring the value of deviation through oral scanning. The mean 3D shoulder and apical and angular deviations were calculated for each group. The data were analyzed using 1-way ANOVA (α=.05 corrected for multiple testing by using Bonferroni-Holm adjustment).
Results:
The mean ±standard deviation 3D shoulder and apical and angular deviations were 0.80 ±0.32 mm, 1.43 ±0.47 mm, and 3.68 ±1.30 degrees. These values were lower than the clinical safety distance of the fully guided implant template. A significantly lower mean 3D apical deviation (1.12 ±0.33 mm, P=.023) and angular deviation (2.81 ±1.11 degrees, P=.018) were observed in group L than in group H (1.68 ±0.37 mm, 4.32 ±0.99 degrees). However, no significant differences were found among the 3 groups in 3D deviation at the shoulder (P>.05).
Conclusions:
AI implant planning software program could design the ideal implant position through self-learning. The accuracy of the AI-assisted designed implant template in this study indicated its clinical reliability. Higher bone density led to increased implant deviations.
