Prediction of strength and strain of the proximal femur by a CT-based finite element method

https://doi.org/10.1016/j.jbiomech.2006.08.003Get rights and content

Abstract

Hip fractures are the most serious complication of osteoporosis and have been recognized as a major public health problem. In elderly persons, hip fractures occur as a result of increased fragility of the proximal femur due to osteoporosis. It is essential to precisely quantify the strength of the proximal femur in order to estimate the fracture risk and plan preventive interventions. CT-based finite element analysis could possibly achieve precise assessment of the strength of the proximal femur. The purpose of this study was to create a simulation model that could accurately predict the strength and surface strains of the proximal femur using a CT-based finite element method and to verify the accuracy of our model by load testing using fresh frozen cadaver specimens. Eleven right femora were collected. The axial CT scans of the proximal femora were obtained with a calibration phantom, from which the 3D finite element models were constructed. Materially nonlinear finite element analyses were performed. The yield and fracture loads were calculated, while the sites where elements failed and the distributions of the principal strains were determined.

The strain gauges were attached to the proximal femoral surfaces. A quasi-static compression test of each femur was conducted. The yield loads, fracture loads and principal strains of the prediction significantly correlated with those measured (r=0.941, 0.979, 0.963). Finite element analysis showed that the solid elements and shell elements in undergoing compressive failure were at the same subcapital region as the experimental fracture site.

Introduction

In elderly persons, hip fractures occur as a result of increased fragility of the proximal femur due to osteoporosis and have been recognized as a major public health problem. Prevention of hip fracture is a high-priority issue because of the rapid increase in the number of elderly people. It is essential to precisely quantify the strength of the proximal femur in order to estimate the fracture risk and plan preventive interventions. Clinically available methods of estimating bone strength include bone densitometry such as dual energy X-ray absorptiometry (DEXA) or peripheral quantitative computed tomography (pQCT), and imaging procedures such as X-ray or CT. These techniques evaluate regional bone density and morphology, which are partly related to fracture risk, but they are of limited value for quantifying structural strength (Faulkner et al., 1993). Therefore, it is necessary to develop a noninvasive method for accurate quantitative structural analysis that incorporates information on both morphology and bone density in a three-dimensional distribution.

CT-based finite element (FE) analysis, which incorporates information on both the three-dimensional architecture and bone density distribution, could possibly achieve precise assessment of the strength of the proximal femur (Cody et al., 1999). Previous studies on CT-based FE analysis of bone strength have shown that it accurately predicts the fracture load (Cody et al., 1999; Keyak, 2001; Keyak et al., 1998). The accuracy of previous methods has not been fully validated, because the studies only evaluated fracture loads and sites, and did not precisely assess strain at the bone surface. Fracture occurs as a result of excessive strain and/or stress within the bone, so it represents the terminal manifestation of the loading process. However, deformation of bone during loading process should also be validated by comparison of results between analytical and experimental data. Therefore, the precision of the CT-based FE method needs to be verified by comparing the strain generated at the bone surface in an experimental setting with the calculated value.

The purpose of this study was to create a simulation model that could accurately predict the strength and surface strains of the proximal femur using a CT-based FE method. The accuracy of our model was verified with comparison of principal strains, displacement, and yield and fracture loads between the prediction and experiment by conducting load testing using fresh frozen cadaver specimens.

Section snippets

Materials and methods

Eleven right femora with no skeletal pathology were collected within 24 hours of death from 5 males aged 30–90 years (average age: 56.8 years) and 6 females aged 52–85 years (average age: 71.5 years). The cause of death was malignant lymphoma (n=3), pneumonia (n=2), myelodysplastic syndrome (n=1), lung cancer (n=1), ovarian cancer (n=1), malignant fibrous histiocytoma (n=1), bladder cancer (n=1) and prostate cancer (n=1). All of the specimens were obtained at the University of Tokyo Hospital

Results

There was a significant linear correlation between the yield load predicted by FE analysis and the measured values (r=0.941, 95% confidence interval, 0.786–0.985; standard error of the estimate SEE=394 N; p<0.0001) (Fig. 6). A significant linear correlation was also noted between the predicted fracture load and the measured values (r=0.979, 95% confidence interval, 0.920–0.995; SEE=228 N; p<0.0001). The slope was 0.936 (not significantly different from 1, p=0.75) with an intercept of 641 N (not

Discussion

The purpose of our study was to create a clinically useful method of predicting the strength of the proximal femur noninvasively in patients with a high fracture risk. We would like to produce a precise model that can predict strength, strain and sites at fracture risk of the proximal femur more accurately than the previous models.

Keyak reported that the slope and the intercept of the regression line for the relation between predicted and experimental fracture loads was 0.77 and 1150 N (SEE: 870 

Acknowledgements

This work was funded by a grant in aid for Scientific Research received from the Japan Society for the Promotion of Science (14657356).

References (33)

  • T.F. Lang et al.

    Volumetric quantitative computed tomography of the proximal femur: precision and relation to bone strength

    Bone

    (1997)
  • M. Lengsfeld et al.

    Comparison of geometry-based and CT voxel-based finite element modelling and experimental validation

    Medical Engineering and Physics

    (1998)
  • D.T. Reilly et al.

    The elastic and ultimate properties of compact bone tissue

    Journal of Biomechanics

    (1975)
  • L. Rohl et al.

    Tensile and compressive properties of cancellous bone

    Journal of Biomechanics

    (1991)
  • D. Ulrich et al.

    Finite element analysis of trabecular bone structure: a comparison of image-based meshing techniques

    Journal of Biomechanics

    (1998)
  • D.G. Altman et al.

    Calculating confidence intervals for regression and correlation

    British Medical Journal (Clinical Research Edition)

    (1988)
  • Cited by (338)

    • Failure analysis of a locking compression plate with asymmetric holes and polyaxial screws

      2023, Journal of the Mechanical Behavior of Biomedical Materials
    View all citing articles on Scopus
    View full text