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

Abstract

Implant peak insertion torque is a commonly used indication of primary stability that dentists rely on to make clinical decisions. The aim of this manuscript is to model the peak torque required for dental implant insertion based on clinical data such as bone properties, implant properties, and drilling procedure. A total of 116 parallel-walled Sweden and Martina dental implants were included in this study. Parameters such as age, sex, bone quality (derived from radiographs), applied peak insertion torque, implant location, implant length, final drill diameter, and implant diameter were recorded. Six data-driven regression models were trained and tested using different combinations of the clinical data to predict the peak torque. A physics-based model was also derived for the peak torque and compared with the data-driven models. The neural network model with early stopping achieved the best accuracy in predicting the clinically measured torque (R² = 0.7692, MSE = 0.08815). Within the limitations of this study, the results suggest that it is possible to predict the peak torque required for implant placement based on the patient's radiographs, implant's properties, and drill diameter. The findings of this study can serve as a reference for dentists in choosing drilling parameters for dental implant surgeries.