Quantifying Accuracy of Self-Tapping Screw Models

Abstract

Correct torquing of bone screws is important to prevent fixation failures and ensure positive patient outcomes. It has been proposed that an automatic model-based method may be able to determine the patient-specific material properties of bone, and provide objective and quantitative torquing recommendations. Models have been previously proposed for identifying the bone material properties, but have not been experimentally tested for accuracy. Here we used these models to perform parameter identification on experimental data using a variety of materials (rigid polyurethane foams) and screws. The identified values were then compared to the values from the datasheet, and matched with a reasonable accuracy for medium-density foam. It was found that for the lower-density foam, the model slightly under-predicted the strength, and for the highest density foam there was a large under-prediction. This suggests that with appropriate calibration, this method is good, but may only be applicable to lower-to-medium strength materials. More thorough testing is required to confirm this and determine the reliable density range.Clinical relevance: Accurate material property identification is required to provide effective torque recommendations for bone screws. This work quantifies the accuracy of two proposed models for material property identification.