Various parameters such as age, height, weight, and body mass index (BMI) influence the hip fracture risk in the elderly which is the most common injury during the sideways fall. This paper presents a parametric study of hip fracture risk based on the gender, age, height, weight, and BMI of subjects using the subject-specific QCT-based finite element modelling and simulation of single-leg stance and sideways fall loadings. Hip fracture risk is estimated using the strain energy failure criterion as a combination of bone stresses and strains leading to more accurate and reasonable results based on the bone failure mechanism. Understanding the effects of various parameters on hip fracture risk can help to prescribe more accurate preventive and treatment plans for a community based on the gender, age, height, weight, and BMI of the population. Results of this study show an increase in hip fracture risk with the increase of age, body height, weight, and BMI in both women and men under the single-leg stance and the sideways fall configurations.
Hip fracture is one of the common injuries during the sideways fall especially for the elderlies where it may cause long-term disability and even death of individuals [
In the past decades, developments and advances in imaging technologies, i.e., Dual-Energy X-ray Absorptiometry (DXA) and Quantitative Computed Tomography (QCT), and numerical methods such as the finite element (FE) method could create a reliable tool for accurate hip fracture risk assessment without the limitations of the statistical models depending on measuring bone mineral density (BMD). A combination of QCT imaging and FE modelling has been employed in many studies to predict hip fracture risk, high stress and strain regions, and failure loads of human femur. For example, Kheirollahi et al. [
In addition to imaging technologies and numerical methods, human body parameters such as age, height, weight, and body mass index (BMI) have great effects on the hip fracture risk assessment. Hence, a study considering various body parameters can be useful to prevent probable hip fracture in the elderly and propose an appropriate treatment for the case of hip fracture occurrence. Parametric studies of hip fracture risk based on the gender, age, height, weight, and BMI of subjects lead to a precise estimation of hip fracture in relation to the specifications and lifestyle of a community. To the authors’ knowledge, there is no study in the literature considering various human body parameters on hip fracture risk using QCT-based FE modelling and strain energy criterion. The main objective of this study is to find the relation of human body parameters with the hip fracture occurrence by integration of QCT imaging technology and FE simulation based on the strain energy criterion which is a combination of both stress and strain effects leading to a more accurate assessment of hip fracture risk [
In this paper, the correlations between the hip fracture risk and different human body parameters such as age, body height, body weight, and BMI are investigated and attained for 30 females (totally 60 right and left femurs) and 30 males (totally 60 right and left femurs). To this purpose, we construct a subject-specific QCT-based FE model of the femur using Mimics and ANSYS software based on the QCT data of subjects and then hip fracture risk index is calculated based on the strain energy criterion using MATLAB codes at the critical regions of the femur, locations usually receiving higher stress and strain within sideways fall [
In this section, the calculation of hip fracture index based on the integration of QCT imaging and FE modelling using the strain energy criterion is explained in detail.
The QCT image is used to construct three dimensional (3D) model of the subject's femur. To this purpose, the QCT images should be saved in the Digital Imaging and Communications in Medicine (DICOM) format and then an appropriate segmentation should be applied to separate the femur from the soft tissue. The intensity of the QCT image voxel is defined as Hounsfield Unit (HU) which is used to determine bone density [
Age (years) | Height (cm) | Body weight (kg) | BMI (kg/m2) | |
---|---|---|---|---|
Range | 50–82 | 149–193.2 | 51.7–126.6 | 18.83–43.36 |
Average | 65 | 169.86 | 81.94 | 28.36 |
To construct the femur 3D model from the subject's QCT image, Mimics software (Materialise, Leuven, Belgium) is employed. QCT image, in DICOM format, is imported to Mimics for the required segmentation and manipulation (
By using the relation between CT numbers and bone material properties, inhomogeneous isotropic material properties of the femur can be obtained from the QCT image data. The bone ash density (
To apply continuous distribution of inhomogeneous bone mechanical properties, elements are categorized into several discrete material bins via Mimics (Materialise, Leuven, Belgium). We need a convergence study to assign the required number of material bins. The maximum von Mises stress is calculated for models with different material bins under the same loading and boundary conditions to find the maximum number of material bins leading to the converged results for use in all FE simulations and numerical analyses.
3D FE model of the femur with the assigned material properties, obtained from Mimics, is imported to ANSYS for further simulations and analyses. Single-leg stance and sideways fall configurations are simulated in the FE analysis. To simulate the single-leg stance, 2.5 times of the subject's body weight, as a distributed load, is applied on the femoral head [
To simulate the sideways fall condition, the distal end of the femur is assumed completely fixed and the femur head is considered constrained in the loading direction (see
Throughout this study, all FE simulations and analyses are done automatically using ANSYS Parametric Design Language (APDL) codes by applying all loading and boundary conditions to a group of nodes at the corresponding locations of the femur (see
Femoral neck, intertrochanteric, and subtrochanteric fractures are three common types of hip fracture (
We define the hip fracture risk index based on the strain energy criterion. The strain energy at the three critical cross-sections of the femur is calculated using MATLAB codes and based on the data obtained from the FE solutions. To generate a two-dimensional (2D) mesh for computing the cross-section strain energy, the plane boundaries of the three critical cross-sections, obtained from the FE mesh, are imported to MATLAB codes. The generated triangle elements over the three critical cross-sections are shown in
The strain energy at the three critical cross-sections of the femur due to the applied forces is obtained as the sum of the strain energy of all triangle elements generated over the cross-section,~i.e.,
Gaussian integration method is employed to compute the strain energy of triangle element
The maximum allowable strain energy or yield strain energy of the three critical cross-sections of the femur is also calculated using in-house MATLAB codes and the data obtained by the APDL codes from the FE solutions. The yield strain energy of the three critical cross-sections is obtained as the sum of the yield strain energy of all triangle elements generated over the cross-section, i.e.,
Hip fracture risk index (FRI) at the three critical cross-sections of the femur using the strain energy criterion is defined as the ratio of the strain energy (
3D FE models of the femur with different material bins are created to investigate model convergence in assigning inhomogeneous material properties. To this purpose, the maximum von Mises stress at the narrowest femoral neck is calculated under the same loading and boundary conditions. As seen from
A convergence study is also conducted to find the element size used in integrating the cross-sectional strain energy. The FRI at the smallest femoral neck cross-section is computed with different maximum element edge lengths. The results are plotted in
To find the required number of integration points for computing the hip FRI, we calculate the FRI at the smallest femoral neck cross-section of 5 clinical cases by considering 3 and 5 integration points (
FRI | |||
---|---|---|---|
Case No. | 3 integration points | 7 integration points | Relative error (%) |
1 | 0.239 | 0.2416 | 1.07 |
2 | 0.6898 | 0.6975 | 1.1 |
3 | 0.2966 | 0.2976 | 0.33 |
4 | 0.8885 | 0.899 | 1.16 |
5 | 1.1482 | 1.1701 | 1.87 |
In this study, the correlations between the hip fracture risk and various human body parameters such as age, body height, weight, and BMI are investigated using 30 females (totally 60 right and left femurs) and 30 males (totally 60 right and left femurs). The correlation coefficients (
Age | Height | Body weight | BMI | |
---|---|---|---|---|
Smallest femoral neck cross-section | −0.0535 |
0.0153 |
0.2933 |
0.3338 |
Intertrochanteric cross-section | 0.0422 |
−0.0805 |
0.2757 |
0.3338 |
Subtrochanteric cross-section | 0.0398 |
−0.0201 |
0.546 |
0.6108 |
Age | Height | Body weight | BMI | |
---|---|---|---|---|
Smallest femoral neck cross-section | −0.0473 |
0.0082 |
0.2133 |
0.2554 |
Intertrochanteric cross-section | 0.0379 |
−0.0576 |
0.0964 |
0.1383 |
Subtrochanteric cross-section | −0.0989 |
0.1332 |
0.1344 |
0.103 |
Age | Height | Body weight | BMI | |
---|---|---|---|---|
Smallest femoral neck cross-section | 0.1464 |
0.0946 |
0.5372 |
0.5038 |
Intertrochanteric cross-section | −0.0358 |
0.2242 |
0.4496 |
0.3788 |
Subtrochanteric cross-section | −0.1778 |
−0.0119 |
0.5681 |
0.6028 |
Age | Height | Body weight | BMI | |
---|---|---|---|---|
Smallest femoral neck cross-section | 0.0697 |
0.2566 |
0.4582 |
0.3504 |
Intertrochanteric cross-section | −0.0475 |
0.4446 |
0.1485 |
−0.0192 |
Subtrochanteric cross-section | 0.0059 |
0.2678 |
0.5059 |
0.4041 |
The FRI variations
The scattered plots of FRI
As seen in
The correlation between FRI and BMI (representative of body shape) is also positive (see
Different human body parameters such as age, height, weight, and BMI influence hip fracture risk in the elderly. Therefore, the correlations of hip fracture risk with age, body height, body weight, and BMI are investigated in this study and the corresponding correlation coefficients are obtained. Generally, there is an increasing trend of hip fracture risk with age, body height, body weight, and BMI at the three critical regions of the femur in both women and men during the single-leg stance and the sideways fall configurations.
The correlations of hip fracture risk with age and height are not significant (
The results of this study also show that to some extent there is a strong correlation between the hip fracture risk and body weight and BMI (
The authors thank the Winnipeg Health Science Centre for providing the QCT images of clinical cases.