Hip Fracture Risk Assessment in Elderly and Diabetic Patients: Combining Autonomous Finite Element Analysis and Machine Learning.

Autonomous finite element analyses (AFE) based on CT scans predict the biomechanical response of femurs during stance and sidewise fall positions. We combine AFE with patient data via a machine learning (ML) algorithm to predict the risk of hip fracture. An opportunistic retrospective clinical study of CT scans is presented, aimed at developing a ML algorithm with AFE for hip fracture risk assessment in type 2 diabetic mellitus (T2DM) and non-T2DM patients. Abdominal/pelvis CT scans of patients who experienced a hip fracture within 2 years after an index CT scan were retrieved from a tertiary medical center database. A control group of patients without a known hip fracture for at least 5 years after an index CT scan was retrieved. Scans belonging to patients with/without T2DM were identified from coded diagnoses. All femurs underwent an AFE under three physiological loads. AFE results, patient's age, weight, and height were input to the ML algorithm (support vector machine [SVM]), trained by 80% of the known fracture outcomes, with cross-validation, and verified by the other 20%. In total, 45% of available abdominal/pelvic CT scans were appropriate for AFE (at least 1/4 of the proximal femur was visible in the scan). The AFE success rate in automatically analyzing CT scans was 91%: 836 femurs we successfully analyzed, and the results were processed by the SVM algorithm. A total of 282 T2DM femurs (118 intact and 164 fractured) and 554 non-T2DM (314 intact and 240 fractured) were identified. Among T2DM patients, the outcome was: Sensitivity 92%, Specificity 88% (cross-validation area under the curve [AUC] 0.92) and for the non-T2DM patients: Sensitivity 83%, Specificity 84% (cross-validation AUC 0.84). Combining AFE data with a ML algorithm provides an unprecedented prediction accuracy for the risk of hip fracture in T2DM and non-T2DM populations. The fully autonomous algorithm can be applied as an opportunistic process for hip fracture risk assessment. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

An approach for simultaneous reduction and fixation of mandibular fractures.

This article presents a new approach for the design of a flexible V-shaped miniplate for mandibular fractures, which combines simultaneous fracture reduction and fixation. A Computerized Tomography (CT) based finite element model was developed to assess the reliability of this design. Muscle and mastication forces were included to replicate post-surgery loading. The V-plate is compared with a standard, linear miniplate, typically used for mandibular fixation. The results indicate that the proposed design can support the fracture while inducing limited fracture displacement, in addition to reducing the duration of the surgery due to fracture reduction by tightening the wire.

Assessing hip fracture risk in type-2 diabetic patients using CT-based autonomous finite element methods : a feasibility study.

AIMS: Type 2 diabetes mellitus (T2DM) impairs bone strength and is a significant risk factor for hip fracture, yet currently there is no reliable tool to assess this risk. Most risk stratification methods rely on bone mineral density, which is not impaired by diabetes, rendering current tests ineffective. CT-based finite element analysis (CTFEA) calculates the mechanical response of bone to load and uses the yield strain, which is reduced in T2DM patients, to measure bone strength. The purpose of this feasibility study was to examine whether CTFEA could be used to assess the hip fracture risk for T2DM patients. METHODS: A retrospective cohort study was undertaken using autonomous CTFEA performed on existing abdominal or pelvic CT data comparing two groups of T2DM patients: a study group of 27 patients who had sustained a hip fracture within the year following the CT scan and a control group of 24 patients who did not have a hip fracture within one year. The main outcome of the CTFEA is a novel measure of hip bone strength termed the Hip Strength Score (HSS). RESULTS: The HSS was significantly lower in the study group (1.76 (SD 0.46)) than in the control group (2.31 (SD 0.74); p = 0.002). A multivariate model showed the odds of having a hip fracture were 17 times greater in patients who had an HSS ≤ 2.2. The CTFEA has a sensitivity of 89%, a specificity of 76%, and an area under the curve of 0.90. CONCLUSION: This preliminary study demonstrates the feasibility of using a CTFEA-based bone strength parameter to assess hip fracture risk in a population of T2DM patients. Cite this article: Bone Joint J 2021;103-B(9):1497-1504.

Patient-specific computed tomography-based finite element analysis: a new tool to assess fracture risk in benign bone lesions of the femur.

BACKGROUND: Most benign active and latent lesions of proximal femur do not predispose a patient to a pathologic fracture. Nonetheless, there is a tendency to perform internal fixation due to the lack of accurate clinical tools that may reliably confirm low risk of pathologic fracture. As many as 30% of these surgeries may be unnecessary. A patient-specific CT-based finite element analysis may quantify bone strength and risk of fracture under normal weight-bearing conditions. METHODS: The clinical relevance of such finite element analysis was investigated in a retrospective study on a cohort of 17 patients. Finite element analysis results (high risk = indication for surgery, low or moderate risk = follow-up) were compared to actual clinical decisions (surgery vs follow-up). All patients predicted by the finite element analysis as high risk underwent internal fixation and had good outcomes (n = 6). FINDINGS: Four of the 11 low- and moderate-risk finite element analysis patients (36%) were operated immediately, and seven (64%) were either operated after a delay of at least 6 months or were never operated. None sustained a pathologic fracture. Patients who were predicted as low fracture risk by finite element analysis remained fracture-free for a minimal period of 6 months. Prediction of high risk of pathologic fracture by finite element analysis was in complete agreement with the conventional clinical evaluation. INTERPRETATION: We consider finite element analysis a promising decision support system for the management of patients with benign tumors of femur, and that it may reliably endorse the decision to withhold surgery for patients at low fracture-risk.

Cold Forming of Al-TiB2 Composites Fabricated by SPS: A Computational Experimental Study.

The mechanical response and failure of Al-TiB2 composites fabricated by Spark Plasma Sintering (SPS) were investigated. The effective flow stress at room temperature for different TiB2 particle volume fractions between 0% and 15% was determined using compression experiments on cylindrical specimens in conjunction with an iterative computational methodology. A different set of experiments on tapered specimens was used to validate the effective flow curves by comparing experimental force-displacement curves and deformation patterns to the ones obtained from the computations. Using a continuum damage mechanics approach, the experiments were also used to construct effective failure curves for each material composition. It was demonstrated that the fracture modes observed in the different experiments could be reproduced in the computations. The results show that increasing the TiB2 particle volume fraction to 10% results in an increase in material effective yield stress and a decrease in hardening. For a particle volume fraction of 15%, the effective yield stress decreases with no significant influence on the hardening slope. The ductility (workability) of the composite decreases with increasing particle volume fraction.

Finite element analyses for predicting anatomical neck fractures in the proximal humerus.

BACKGROUND: Proximal humerus fractures which occur as a result of a fall on an outstretched arm are frequent among the elderly population. The necessity of stabilizing such fractures by surgical procedures is a controversial matter among surgeons. Validating a personalized FE analysis by ex-vivo experiments of humeri and mimicking such fractures by experiments is the first step along the path to determine the necessity of such surgeries. METHODS: Four fresh frozen human humeri were loaded using a new simple experimental setting, so to fracture the humeri at the anatomical neck. Strains on humeri's surfaces predicted by the high order FE analyses (as in Dahan et al., 2016) were compared to the experimental observations to further enhance the validity of the FE analyses. A simplified yield criterion based on a linear elastic analysis and principal strains was used to predict the anatomical neck fracture as observed in the experiment. FINDINGS: An excellent correlation between experimental measured and FE predicted strains was obtained (slope of 0.99 and R2=0.98). All humeri were fractured at the anatomical neck. The predicted yield load was within 10%-20% accuracy. INTERPRETATION: High-order FE analyses reliably predict strains and yield loads in the humeri. Fractures induced by the experimental setting correspond to anatomical neck fractures noticed in practice and classified as AO C1.1-C1.3. Surgical neck fractures, which are most common in clinical practice, could not be realized in the proposed experiments, and a different experimental setting should be sought to obtain them ex-vivo.

Pathological fracture risk assessment in patients with femoral metastases using CT-based finite element methods. A retrospective clinical study.

Physician recommendation for prophylactic surgical fixation of a femur with metastatic bone disease (MBD) is usually based on Mirels' criteria and clinical experience, both of which suffer from poor specificity. This may result in a significant number of these health compromised patients undergoing unnecessary surgery. CT-based finite element analyses (CTFEA) have been shown to accurately predict strength in femurs with metastatic tumors in an ex-vivo study. In order to assess the utility of CTFEA as a clinical tool to determine the need for fixation of patients with MBD of the femur, an ad hoc CTFEA was performed on a retrospective cohort of fifty patients. Patients with CT scans appropriate for CTFEA analysis were analyzed. Group 1 was composed of 5 MBD patients who presented with a pathologic femoral fracture and had a scan of their femurs just prior to fracture. Group 2 was composed of 45 MBD patients who were scheduled for a prophylactic surgery because of an impending femoral fracture. CTFEA models were constructed for both femurs for all patients, loaded with a hip contact force representing stance position loading accounting for the patient's weight and femur anatomy. CTFEA analysis of Group 1 patients revealed that they all had higher tumor associated strains compared to typical non-diseased femur bone strains at the same region (>45%). Based on analysis of the 5 patients in Group 1, the ratio between the absolute maximum principal strain in the vicinity of the tumor and the typical median strain in the region of the tumor of healthy bones (typical strain fold ratio) was found to be the 1.48. This was considered to be the predictive threshold for a pathological femoral fracture. Based on this typical strain fold ratio, twenty patients (44.4%) in Group 2 were at low risk of fracture and twenty-five patients (55.5%) high risk of fracture. Eleven patients in Group 2 choose not to have surgery and none fractured in the 5month follow-up period. CTFEA predicted that seven of these patients were below the pathological fracture threshold and four above, for a specificity of 63% Based on CTFEA, 39% of the patients with femoral MBD who were referred and underwent prophylactic stabilization may not have needed surgery. These results indicate that a prospective randomized clinical trial evaluating CTFEA as a criterion for determining the need for surgical stabilization in patients with MBD of the femur may be warranted.

Verified and validated finite element analyses of humeri.

BACKGROUND: Although ~200,000 emergency room visits per year in the US alone are associated with fractures of the proximal humerus, only limited studies exist on their mechanical response. We hypothesise that for the proximal humeri (a) the mechanical response can be well predicted by using inhomogeneous isotropic material properties, (b) the relation between bone elastic modulus and ash density (E(ρash)) is similar for the humerus and the femur, and may be general for long bones, and (c) it is possible to replicate a proximal humerus fracture in vitro by applying uniaxial compression on humerus׳ head at a prescribed angle. METHODS: Four fresh frozen proximal humeri were CT-scanned, instrumented by strain-gauges and loaded at three inclination angles. Thereafter head displacement was applied to obtain a fracture. CT-based high order (p-) finite element (FE) and classical (h-) FE analyses were performed that mimic the experiments and predicted strains were compared to the experimental observations. RESULTS: The E(ρash) relationship appropriate for the femur is equally appropriate for the humeri: predicted strains in the elastic range showed an excellent agreement with experimental observations with a linear regression slope of m=1.09 and a coefficient of regression R(2)=0.98. p-FE and h-FE results were similar for the linear elastic response. Although fractures of the proximal humeri were realised in the in vitro experiments, the contact FE analyses (FEA) were unsuccessful in representing properly the experimental boundary conditions. DISCUSSION: The three hypotheses were confirmed and the linear elastic response of the proximal humerus, attributed to a stage at which the cortex bone is intact, was well predicted by the FEA. Due to a large post-elastic behaviour following the cortex fracture, a new non-linear constitutive model for proximal humerus needs to be incorporated into the FEA to well represent proximal humerus fractures. Thereafter, more in vitro experiments are to be performed, under boundary conditions that may be well represented by the FEA, to allow a reliable simulation of the fracture process.

Predicting the stiffness and strength of human femurs with real metastatic tumors.

BACKGROUND: Predicting patient specific risk of fracture in femurs with metastatic tumors and the need for surgical intervention are of major clinical importance. Recent patient-specific high-order finite element methods (p-FEMs) based on CT-scans demonstrated accurate results for healthy femurs, so that their application to metastatic affected femurs is considered herein. METHODS: Radiographs of fresh frozen proximal femur specimens from donors that died of cancer were examined, and seven pairs with metastatic tumor were identified. These were CT-scanned, instrumented by strain-gauges and loaded in stance position at three inclination angles. Finally the femurs were loaded until fracture that usually occurred at the neck. Histopathology was performed to determine whether metastatic tumors are present at fractured surfaces. Following each experiment p-FE models were created based on the CT-scans mimicking the mechanical experiments. The predicted displacements, strains and yield loads were compared to experimental observations. RESULTS: The predicted strains and displacements showed an excellent agreement with the experimental observations with a linear regression slope of 0.95 and a coefficient of regression R(2)=0.967. A good correlation was obtained between the predicted yield load and the experimental observed yield, with a linear regression slope of 0.80 and a coefficient of regression R(2)=0.78. DISCUSSION: CT-based patient-specific p-FE models of femurs with real metastatic tumors were demonstrated to predict the mechanical response very well. A simplified yield criterion based on the computation of principal strains was also demonstrated to predict the yield force in most of the cases, especially for femurs that failed at small loads. In view of the limited capabilities to predict risk of fracture in femurs with metastatic tumors used nowadays, the p-FE methodology validated herein may be very valuable in making clinical decisions.

Patient-specific FE analyses of metatarsal bones with inhomogeneous isotropic material properties.

The mechanical response of human metatarsal bones is of importance in both research and clinical practice, especially when associated with the correction of Hallux Valgus. Verified and validated patient-specific finite-element analysis (FEA) based on CT scans developed for human femurs are extended here to the first and second metatarsal bones. Two fresh-frozen metatarsal #1 and five metatarsal #2 bones from three donors were loaded in-vitro at three different angles. Holes typical to Hallux Valgus correction were then drilled in the bones, which were reloaded until fracture. In parallel, high-order FE models of the bones were created from CT-scans that mimic the experimental setting. We validated the FE results by comparison to experimental observations. Excellent agreement was obtained with R(2)=0.97 and slope of the regression line close to 1. We also compared the FE predicted fracture load and location for the second metatarsal bones with these measured in the experiment, demonstrating an excellent prediction within 10% difference. After validation of the FE predictions, they were used to investigate the effect of drilled hole position, dimension and the insertion of a metallic device on the mechanical response so to optimize the outcome of the Hallux Valgus correction. This study further substantiates the potential use of FEA in clinical practice.

The finite cell method for bone simulations: verification and validation.

Standard methods for predicting bone's mechanical response from quantitative computer tomography (qCT) scans are mainly based on classical h-version finite element methods (FEMs). Due to the low-order polynomial approximation, the need for segmentation and the simplified approach to assign a constant material property to each element in h-FE models, these often compromise the accuracy and efficiency of h-FE solutions. Herein, a non-standard method, the finite cell method (FCM), is proposed for predicting the mechanical response of the human femur. The FCM is free of the above limitations associated with h-FEMs and is orders of magnitude more efficient, allowing its use in the setting of computational steering. This non-standard method applies a fictitious domain approach to simplify the modeling of a complex bone geometry obtained directly from a qCT scan and takes into consideration easily the heterogeneous material distribution of the various bone regions of the femur. The fundamental principles and properties of the FCM are briefly described in relation to bone analysis, providing a theoretical basis for the comparison with the p-FEM as a reference analysis and simulation method of high quality. Both p-FEM and FCM results are validated by comparison with an in vitro experiment on a fresh-frozen femur.

Patient-specific finite-element analyses of the proximal femur with orthotropic material properties validated by experiments.

Patient-specific high order finite-element (FE) models of human femurs based on quantitative computer tomography (QCT) with inhomogeneous orthotropic and isotropic material properties are addressed. The point-wise orthotropic properties are determined by a micromechanics (MM) based approach in conjunction with experimental observations at the osteon level, and two methods for determining the material trajectories are proposed (along organs outer surface, or along principal strains). QCT scans on four fresh-frozen human femurs were performed and high-order FE models were generated with either inhomogeneous MM-based orthotropic or empirically determined isotropic properties. In vitro experiments were conducted on the femurs by applying a simple stance position load on their head, recording strains on femurs' surface and head's displacements. After verifying the FE linear elastic analyses that mimic the experimental setting for numerical accuracy, we compared the FE results to the experimental observations to identify the influence of material properties on models' predictions. The strains and displacements computed by FE models having MM-based inhomogeneous orthotropic properties match the FE-results having empirically based isotropic properties well, and both are in close agreement with the experimental results. When only the strains in the femoral neck are being compared a more pronounced difference is noticed between the isotropic and orthotropic FE result. These results lay the foundation for applying more realistic inhomogeneous orthotropic material properties in FEA of femurs.

Patient-specific finite element analysis of the human femur--a double-blinded biomechanical validation.

Patient-specific finite element (PSFE) models based on quantitative computer tomography (qCT) are generally used to "predict" the biomechanical response of human bones with the future goal to be applied in clinical decision-making. However, clinical applications require a well validated tool that is free of numerical errors and furthermore match closely experimental findings. In previous studies, not all measurable data (strains and displacements) were considered for validation. Furthermore, the same research group performed both the experiments and PSFE analyses; thus, the validation may have been biased. The aim of the present study was therefore to validate PSFE models with biomechanical experiments, and to address the above-mentioned issues of measurable data and validation bias. A PSFE model (p-method) of each cadaver femur (n = 12) was generated based on qCT scans of the specimens. The models were validated by biomechanical in-vitro experiments, which determined strains and local displacements on the bone surface and the axial stiffness of the specimens. The validation was performed in a double-blinded manner by two different research institutes to avoid any bias. Inspecting all measurements (155 values), the numerical results correlated well with the experimental results (R(2) = 0.93, slope 1.0093, mean of absolute deviations 22%). In conclusion, a method to generate PSFE models from qCT scans was used in this study on a sample size not yet considered in the past, and compared to experiments in a douple-blinded manner. The results demonstrate that the presented method is in an advanced stage, and can be used in clinical computer-aided decision-making.

Predicting the yield of the proximal femur using high-order finite-element analysis with inhomogeneous orthotropic material properties.

High-order finite-element (FE) analyses with inhomogeneous isotropic material properties have been shown to predict the strains and displacements on the surface of the proximal femur with high accuracy when compared with in vitro experiments. The same FE models with inhomogeneous orthotropic material properties produce results similar to those obtained with isotropic material properties. Herein, we investigate the yield prediction capabilities of these models using four different yield criteria, and the spread in the predicted load between the isotropic and orthotropic material models. Subject-specific high-order FE models of two human femurs were generated from CT scans with inhomogeneous orthotropic or isotropic material properties, and loaded by a simple compression force at the head. Computed strains and stresses by both the orthotropic and isotropic FE models were used to determine the load that predicts 'yielding' by four different 'yield criteria': von Mises, Drucker-Prager, maximum principal stress and maximum principal strain. One of the femurs was loaded by a simple load until fracture, and the force resulting in yielding was compared with the FE predicted force. The surface average of the 'maximum principal strain' criterion in conjunction with the orthotropic FE model best predicts both the yield force and fracture location compared with other criteria. There is a non-negligible influence on the predictions if orthotropic or isotropic material properties are applied to the FE model. All stress-based investigated 'yield criteria' have a small spread in the predicted failure. Because only one experiment was performed with a rather simplified loading configuration, the conclusions of this work cannot be claimed to be either reliable or sufficient, and future experiments should be performed to further substantiate the conclusions.

Validation of subject-specific automated p-FE analysis of the proximal femur.

BACKGROUND: The use of subject-specific finite element (FE) models in clinical practice requires a high level of automation and validation. In Yosibash et al. [2007a. Reliable simulations of the human proximal femur by high-order finite element analysis validated by experimental observations. J. Biomechanics 40, 3688-3699] a novel method for generating high-order finite element (p-FE) models from CT scans was presented and validated by experimental observations on two fresh frozen femurs (harvested from a 30 year old male and 21 year old female). Herein, we substantiate the validation process by enlarging the experimental database (54 year old female femur), improving the method and examine its robustness under different CT scan conditions. APPROACH: A fresh frozen femur of a 54 year old female was scanned under two different environments: in air and immersed in water (dry and wet CT). Thereafter, the proximal femur was quasi-statically loaded in vitro by a 1000N load. The two QCT scans were manipulated to generate p-FE models that mimic the experimental conditions. We compared p-FE displacements and strains of the wet CT model to the dry CT model and to the experimental results. In addition, the material assignment strategy was reinvestigated. The inhomogeneous Young's modulus was represented in the FE model using two different methods, directly extracted from the CT data and using continuous spatial functions as in Yosibash et al. [2007a. Reliable simulations of the human proximal femur by high-order finite element analysis validated by experimental observations. J. Biomechanics 40, 3688-3699]. RESULTS: Excellent agreement between dry and wet FE models was found for both displacements and strains, i.e. the method is insensitive to CT conditions and may be used in vivo. Good agreement was also found between FE results and experimental observations. The spatial functions representing Young's modulus are local and do not influence strains and displacements prediction. Finally, the p-FE results of all three fresh frozen human femurs compare very well to experimental observations exemplifying that the presented method may be in a mature stage to be used in clinical computer-aided decision making.

Reliable simulations of the human proximal femur by high-order finite element analysis validated by experimental observations.

BACKGROUND: The mechanical response of patient-specific bone to various load conditions is of major clinical importance in orthopedics. Herein we enhance the methods presented in Yosibash et al. [2007. A CT-based high-order finite element analysis of the human proximal femur compared to in-vitro experiments. ASME Journal of Biomechanical Engineering 129(3), 297-309.] for the reliable simulations of the human proximal femur by high-order finite elements (FEs) and validate the simulations by experimental observations. METHOD OF APPROACH: A fresh-frozen human femur was scanned by quantitative computed tomography (QCT) and thereafter loaded (in vitro experiments) by a quasi-static force of up to 1250 N. QCT scans were manipulated to generate a high-order FE bone model with distinct cortical and trabecular regions having inhomogeneous isotropic elastic properties with Young's modulus represented by continuous spatial functions. Sensitivity analyses were performed to quantify parameters that mostly influence the mechanical response. FE results were compared to displacements and strains measured in the experiments. RESULTS: Young moduli correlated to QCT Hounsfield Units by relations in Keyak and Falkinstein [2003. Comparison of in situ and in vitro CT scan-based finite element model predictions of proximal femoral fracture load. Medical Engineering and Physics 25, 781-787.] were found to provide predictions that match the experimental results closely. Excellent agreement was found for both the displacements and strains. The presented study demonstrates that reliable and validated high-order patient-specific FE simulations of human femurs based on QCT data are achievable for clinical computer-aided decision making.