Development of a peak insertion torque prediction model for parallel-walled dental implants

Area of Science:

• Biomaterials Engineering
• Dental Implantology
• Biomechanical Engineering

Background:

• Implant peak insertion torque is a critical indicator of primary stability in dental implantology.
• Dentists utilize insertion torque to guide clinical decisions during implant placement procedures.

Purpose of the Study:

• To develop predictive models for dental implant insertion torque.
• To investigate the influence of bone and implant properties, and drilling parameters on insertion torque.
• To compare data-driven models with a physics-based model for torque prediction.

Main Methods:

• Collected clinical data from 116 dental implant placements, including patient demographics, bone quality from radiographs, implant characteristics, and drilling parameters.
• Trained and tested six data-driven regression models, including a neural network with early stopping.
• Developed and compared a physics-based model against the data-driven models.

Main Results:

• The neural network model demonstrated the highest accuracy in predicting clinically measured torque (R² = 0.7692, MSE = 0.08815).
• Predictive accuracy was achieved using patient radiographs, implant properties, and drill diameter.
• The study successfully modeled peak insertion torque based on clinical and procedural variables.

Conclusions:

• It is feasible to predict the peak insertion torque for dental implants using readily available clinical data.
• The findings provide a valuable reference for dentists to select optimal drilling parameters.
• Accurate torque prediction can enhance primary stability and inform clinical decision-making in implant dentistry.