Predicting overloading plate failure using specimen-specific finite element models combined with implantable sensors

Area of Science:

• Biomedical Engineering
• Orthopedic Surgery
• Biomechanics

Background:

• Mechanical failures in plate osteosynthesis, like plate bending, remain a clinical challenge.
• Finite element (FE) models can simulate bone-plate constructs but lack in vivo validation due to unknown loads.
• Implantable sensors offer a path to validate FE models by measuring in vivo plate deformation during fracture healing.

Purpose of the Study:

• To bridge the gap between FE simulations and sensor data for predicting implant failure.
• To establish a link between sensor signals and predicted mechanical failure in bone-plate constructs.
• To enable retrospective in vivo validation of FE models using implantable sensor data.

Main Methods:

• Seven ovine tibia fractures were fixed with locking plates and tested quasi-statically for failure.
• Implantable sensors monitored plate bending deformation during failure testing.
• FE models of the constructs were created, incorporating calibrated virtual sensor signals.

Main Results:

• High correlation (R² > 0.99) was observed between experimental and virtual sensor signals in initial tests.
• FE models showed strong correlation (concordance correlation coefficient = 0.89) with experimental sensor signals at yield.
• Virtual sensor signals accurately predicted experimental signals at the onset of plate bending.

Conclusions:

• FE models can accurately predict sensor signals at plate bending onset, validating models retrospectively.
• This approach allows for in vivo validation without requiring direct load measurements.
• Findings support the development of tailored rehabilitation strategies to reduce patient complications.