Work

Evaluation Using AI-Based Approach of Pedicle Screw Stability Using Laser Resonance Frequency Analysis Prior to Surgical Insertion

Year: 2026

Type: article

Abstract: In orthopedics, assessing the initial stability of pedicle screws is crucial for preventing failure after implantation. Traditional methods for determining stability are applicable only to post-screw insertion, complicating recovery in cases of inadequate strength. This study aimed to predict implant stability at the tapping stage prior to screw placement. Artificial bones with five distinct densities were utilized to measure tapping and screw insertion torques, facilitating a statistical evaluation of bone density and placement strength. A laser resonance frequency analysis (L-RFA) was conducted to obtain the vibration frequency spectra of the tap. Furthermore, linear approximation and support vector regression (SVR) were employed, utilizing spectral data and a 95% confidence interval to predict bone density. Significant differences in tapping and screw insertion torques were observed relative to bone density ( p <0.01). Prediction based on a wide frequency range utilizing SVR improved the mean squared error from 26.9 to 8.08 when compared to traditional regression prediction, resulting in a coefficient of determination R 2 of 0.843. The acquisition of bone density data at the tapping stage enables the prediction of expected screw stability. Bone density prediction using L-RFA has the potential to provide stability indicators prior to pedicle screw insertion.

Source: Journal of Advanced Computational Intelligence and Intelligent Informatics

DOI: 10.20965/jaciii.2026.p0718