New insights on the proximal femur biomechanics using Digital Image Correlation
Introduction
Finite element (FE) models of femurs based on quantitative computed tomography (QCT) have been used extensively to estimate bone stiffness and strength (Yosibash et al., 2007, Trabelsi et al., 2011, Engelke et al., 2016, de Bakker et al., 2017). These models may be used in clinical practice for fracture risk prediction due to osteoporosis and metastases (Yosibash et al., 2014, Dragomir-Daescu et al., 2011, Keyak et al., 2011, Sternheim et al., 2018) as well as for an optimal implant selection (Katz et al., 2018, Cilla et al., 2017). FE bone models were usually validated by comparing computed surface strains to experimental measurements at same locations. Average strains on a small surface, measured by strain gauges (SGs) are considered to be the gold standard in FE model validation of femurs (Trabelsi et al., 2011, Wille and Rank, 2012, Eberle et al., 2013, Pettersen et al., 2009, Ruess et al., 2012, Schileo et al., 2014). In recent years, digital image correlation (DIC) has emerged as a promising technology to validate FE models accounting for thousands of measured points (Jetté et al., 2017, Grassi et al., 2016, Grassi et al., 2014, Grassi et al., 2013, Väänänen et al., 2013, Dickinson et al., 2011). DIC is a non-contact optical technique that enables a three-dimensional (3D), full field measurement and a complete evaluation of displacements and strains on bone surface.
Here, we apply DIC techniques on cadaveric femurs to monitor their mechanical response at areas where highest principal strains are noticed including the femoral neck which has been mostly neglected in past publications. These measurements are used to further enhance the validation of QCT based FE models, providing new insights on the mechanical behavior and modelling of the femoral neck, which is of high importance for fracture prediction.
Only few validation studies performed 3D DIC measurements on femurs loaded in a stance position, composite femurs were mostly considered (Jetté et al., 2017, Väänänen et al., 2013, Grassi et al., 2013, Dickinson et al., 2011). DIC measurements are easier on composite femurs because of their smooth and dry surface, whereas cadaveric femurs have rough surfaces in some areas and are wet or greasy (which may cause glare). Additionally, composite femurs do not represent well the geometry and material properties of cadaver femurs (Grant et al., 2007, Zdero et al., 2008). To the best of our knowledge there exists only one study that addressed actual femurs (Grassi et al., 2014), however the superior neck region was not reported nor a validation of the DIC to gold standard SGs was performed. Such a comparison between DIC and SG was not yet documented on cadaver femurs (Ghosh et al., 2012, Gilchrist et al., 2013).
Femoral frontal surface is most commonly monitored by DIC (Väänänen et al., 2013, Grassi et al., 2013, Grassi et al., 2014, Grassi et al., 2016, Jetté et al., 2017). This approach has some limitations: (a) Because strains are derivatives of the displacements, DIC technique cannot calculate strains on the edges of the viewed region where the extreme values are found. Also, strains on these edges may be erroneous due to the curvature of the specimen (Jetté et al., 2017, Dickinson et al., 2011). (b) Area of measurement is close to the neutral axis where strains are low and noise may become dominant.
Although the superior neck is of special interest since neck fractures usually start there in stance position experiments, no DIC measurements are reported in past publications for that location. Displacements measured by DIC are also rarely addressed. The single study to our knowledge accounting for displacements is by Jetté et al. (2017), where composite femurs were used.
Here five fresh frozen femurs were used to validate QCT based high order FE models (one of the femurs has sever osteoarthritis as may be common in the elderly population). DIC strains were compared to strain gauge measurements and thereafter compared to FE computed strains for validation. Displacements were also addressed for validation. Lateral and medial surfaces of shaft and neck were examined (similarly to past studies performed with SGs). These surfaces experience highest strains and thus are of interest. DIC measurements on the neck’s superior surface allow new insights on its mechanical behaviour and modelling.
Section snippets
Methods
Five fresh frozen femurs were defrosted, thoroughly cleaned of soft tissue using scalpels and tweezers, cut at the shaft 240 mm distally of the lesser trochanter and imbedded inside a steel cylinder using PMMA (Fig. 1). Femurs were CT scanned with calibration phantoms (Mindways, 2002, Katz et al., 2019). Donors’ and CT details are given in Table 1.
Femurs were painted white and speckled black using spray paint (white matte aerosol spray by Tambour LTD and black 2X Ultra Cover Flat Spray).
Comparing DIC to SG
A comparison between SG strains to DIC strains is presented in Fig. 3. Locations of the SGs inside the analyzed strain field are given in the supplementary material (Fig. A.1).
RMSEs between DIC and SG were below 50 μs in most cases. A poor agreement was obtained for SG2 at FFI1L neck (RMSE = 407 μs). This SG was located near the edge of the DIC analyzed field which lead to high noise in the DIC strains. Because of the noise, principal strain is not zero at zero load after time smoothening.
Discussion
To the best of our knowledge, this is the first study that considered full field measurements using two DIC systems (4 cameras) simultaneously to monitor the mechanical response of fresh frozen cadaver femurs, monitoring medial and lateral bone surfaces which experience the highest strains. In previous studies (Jetté et al., 2017, Grassi et al., 2013, Grassi et al., 2016, Väänänen et al., 2013) the femoral frontal surface (including neutral axis) was monitored. In these studies, the superior
Limitations and future work
The observations reported in the current study are based on five femurs, only in four of which the superior neck was monitored (three different donors). Further specimens should be investigated in the future to further increase the sample size and enhance validity of the conclusions.
The misrepresentation of the femoral neck mechanical response demand further investigation. The high strains at the neck ’saddle region’ may be due to poor cortex identification in the CT scan (Prevrhal et al., 2003
Declaration of Competing Interest
YK has no conflicts of interest. ZY has a financial interest in PerSimiO.
Acknowledgements
The authors thank Dr. Nimrod Snir from Sourasky Medical Center, Tel Aviv, Israel for his help in cleaning the bones from soft tissue and helping with the CT-scans and Ms. Gal Dahan from TAU for her help with the experiments.
References (40)
- et al.
Individual density - elasticity relationships improve accuracy of subject-specific finite element models of human femurs
J. Biomech.
(2013) - et al.
Artificial composite bone as a model of human trabecular bone: the implant bone interface
J. Biomech.
(2007) - et al.
How accurately can subject-specific finite element models predict strains and strength of human femora? Investigation using full-field measurements
J. Biomech.
(2016) - et al.
Experimental validation of finite element model for proximal composite femur using optical measurements
J. Mech. Behav. Biomed. Mater.
(2013) - et al.
Development of a balanced experimental-computational approach to understanding the mechanics of proximal femur fractures
Med. Eng. Phys.
(2014) - et al.
The human proximal femur behaves linearly elastic up to failure under physiological loading conditions
J. Biomech.
(2011) - et al.
Scanner influence on the mechanical response of QCT-based finite element analysis of long bones
J. Biomech.
(2019) - et al.
Patient-specific finite element analysis of femurs with cemented hip implants
Clin. Biomech.
(2018) - et al.
Prediction of fracture location in the proximal femur using finite element models
Med. Eng. Phys.
(2001) - et al.
Male - female differences in the association between incident hip fracture and proximal femoral strength: a finite element analysis study
Bone
(2011)
To what extent can linear finite element models of human femora predict failure under stance and fall loading configurations?
J. Biomech.
An accurate estimation of bone density improves the accuracy of subject-specific finite element models
J. Biomech.
Pathological fracture risk assessment in patients with femoral metastases using ct-based finite element methods. a retrospective clinical study
Bone
Patient-specific finite element analysis of the human femur - a double-blinded biomechanical validation
J. Biomech.
Repeatability of digital image correlation for measurement of surface strains in composite long bones
J. Biomech.
Prediction of the mechanical response of the femur with uncertain elastic properties
J. Biomech.
Predicting the stiffness and strength of human femurs with real metastatic tumors
Bone
Reliable simulations of the human proximal femur by high-order finite element analysis validated by experimental observations
J. Biomech.
On the failure initiation in the proximal human femur under simulated sideways fall
Ann. Biomed. Eng.
Clinical evaluation of bone strength and fracture risk
Curr. Osteoporos. Rep.
Cited by (26)
Femurs segmentation by machine learning from CT scans combined with autonomous finite elements in orthopedic and endocrinology applications
2023, Computers and Mathematics with ApplicationsThe predictive ability of a QCT-FE model of the proximal femoral stiffness under multiple load cases is strongly influenced by experimental uncertainties
2023, Journal of the Mechanical Behavior of Biomedical MaterialsDigital image correlation based on convolutional neural networks
2023, Optics and Lasers in EngineeringCitation Excerpt :Over the past three decades, digital image correlation (DIC) technique has already become a classical non-destructive photomechanical method to characterize spatio-temporal deformation fields of specimen surfaces [1–6]. Owing to its flexibility, reliability, robustness and ease of use [7–9], DIC has been increasingly applied to a large number of engineering kingdoms concerning non-contact full-field measurements, such as experimental mechanics [10–14], biomechanics/cell mechanics [15–17], structural health monitoring [18,19], fracture mechanics [20–22] and composite mechanics [23–26]. The core of the traditional DIC algorithms lies in implementing numerical comparisons between undeformed speckle images (i.e., reference images) and the corresponding deformed ones (i.e., target images) and eventually acquiring full-field displacement and strain knowledges [27–31].
Role of impaction bone grafting of allografts in the management of benign lesions of the proximal femur
2022, Journal of OrthopaedicsThe influence of foramina on femoral neck fractures and strains predicted with finite element analysis
2022, Journal of the Mechanical Behavior of Biomedical MaterialsCitation Excerpt :The larger region where high strains are measured, compared to the predicted strains, suggests that other factors, such as remaining soft tissue, could play a role. Similar results have been observed under loading in single-leg-stance, where high strains measured with DIC were observed outside of the foramina (Katz and Yosibash, 2020). These strains were also not captured by FE models based on clinical CT scans.
A credible homogenized finite element model to predict radius fracture in the case of a forward fall
2022, Journal of the Mechanical Behavior of Biomedical Materials