Pull out strength calculator for pedicle screws using a surrogate ensemble approach

Area of Science:

• Spinal surgery
• Biomechanical engineering
• Medical device analysis

Background:

• Pedicle screw instrumentation is crucial for treating spinal disorders.
• Current methods for assessing screw holding power are subjective, relying on surgeon's feel.
• There is a need for objective, predictive tools for pedicle screw performance.

Purpose of the Study:

• To develop a predictive surrogate model for pedicle screw pullout strength.
• To identify key factors influencing pedicle screw holding power.
• To enhance pre-surgical planning and decision support for spinal surgeons.

Main Methods:

• Experimental design using Taguchi's orthogonal array to identify influential factors.
• Pullout strength testing of polyaxial pedicle screws on polyurethane foam blocks (ASTM F543).
• Development and evaluation of surrogate models (Krigging, polynomial response surface, radial basis function, and ensemble weighted average).

Main Results:

• Density, insertion depth, insertion angle, and reinsertion significantly affect pedicle screw pullout strength (p < 0.05).
• The weighted average surrogate model demonstrated superior predictive accuracy for pullout strength.
• The developed model provides a reliable prediction of screw holding power.

Conclusions:

• A robust predictive model for pedicle screw pullout strength has been successfully developed.
• This model integrates experimental data with advanced surrogate modeling techniques.
• The predictive tool can significantly aid spine surgeons in pre-surgical planning and decision-making.