A new method to determine cortical bone thickness in CT images using a hybrid approach of parametric profile representation and local adaptive thresholds: Accuracy results

Can neck fractures in proximal humeri be predicted by CT-based FEA?

BACKGROUND

Proximal humeri fractures at anatomical and surgical neck (∼5% and ∼50% incidence respectively) are frequent in elderly population. Yet, neither in-vitro experiments nor CT-based finite element analyses (CTFEA) have investigated these in depth. Herein we enhance (Dahan et al., 2019) (addressing anatomical neck fractures) by more experiments and specimens, accounting for surgical neck fractures and explore CTFEA's prediction of humeri mechanical response and yield force.

METHODS

Four fresh frozen human humeri were tested in a new experimental configuration inducing surgical neck fractures. Digital image correlation (DIC) provided strains and displacements on humeri surfaces and used to validate CTFEA predictions. CTFEA were enhanced herein to improve the accuracy at the proximal neck: A cortical bone mapping (CBM) algorithm was implemented to overcome insufficient scanning resolution, and a new trabecular material mapping was investigated.

RESULTS

The new experimental setting induced impacted surgical neck fractures in all humeri. Excellent DIC to CTFEA correlation in strains was obtained at the shaft (slope 0.984, R²=0.99) and a fair agreement (slope 0.807, R²=0.73) at the neck. CBM algorithm had worsened the correlation, whereas the new material mapping had a negligible influence. Yield loads predictions improved considerably when trabecular yielding (maximum principal strain criterion) was considered instead of surface cortical yielding.

DISCUSSION

CTFEA well predicts strains on the shaft and reasonably well on the neck. This enhances former conclusions by past studies conducted using SGs, now also evident by DIC. Yield load prediction for surgical neck fractures (involving crushing of trabecular bone) is predicted better by trabecular failure laws rather than cortex ones. Further FEA studies using trabecular orthotropic constitutive models and failure laws are warrant.

Focal osteoporosis defect is associated with vertebral compression fracture prevalence in a bone mineral density-independent manner

Introduction

Focal osteoporosis defect has shown a high association with the bone fragility and osteoporotic fracture prevalence. However, no routine computed tomography (CT)-based vertebral focal osteoporosis defect measurement and its association with vertebral compression fracture (VCF) were discussed yet. This study aimed to develop a routine CT-based measurement method for focal osteoporosis defect quantification, and to assess its association with the VCF prevalence.

Materials and Methods

A total of 205 cases who underwent routine CT scanning, were retrospectively reviewed and enrolled into either the VCF or the control group. The focal bone mineral content loss (focal BMC loss), measured as the cumulated demineralization within bone void space, was proposed for focal osteoporosis defect quantification. Its scan-rescan reproducibility and its correlation with trabecular bone mineral density (BMD) and apparent microarchitecture parameters were evaluated. The association between focal BMC loss and the prevalence of VCF was studied by logistic regression.

Results

The measurement of focal BMC loss showed high reproducibility (RMSSD = 0.011 mm, LSC = 0.030 mm, ICC = 0.97), and good correlation with focal bone volume fraction (r = 0.79, P < 0.001), trabecular bone separation (r = 0.76, P < 0.001), but poor correlation with trabecular BMD (r = 0.37, P < 0.001). The focal BMC loss was significantly higher in the fracture group than the control (1.03 ± 0.13 vs. 0.93 ± 0.11 mm; P < 0.001), and was associated with prevalent VCF (1.87, 95% CI = 1.31-2.65, P < 0.001) independent of trabecular BMD level.

Discussion

As a surrogate measure of focal osteoporosis defect, focal BMC Loss independently associated with the VCF prevalence. It suggests that focal osteoporosis defect is a common manifestation that positively contributed to compression fracture risk and can be quantified with routine CT using focal BMC Loss.

Automated cortical thickness measurement of the mandibular condyle head on CBCT images using a deep learning method

This study proposes a deep learning model for cortical bone segmentation in the mandibular condyle head using cone-beam computed tomography (CBCT) and an automated method for measuring cortical thickness with a color display based on the segmentation results. In total, 12,800 CBCT images from 25 normal subjects, manually labeled by an oral radiologist, served as the gold-standard. The segmentation model combined a modified U-Net and a convolutional neural network for target region classification. Model performance was evaluated using intersection over union (IoU) and the Hausdorff distance in comparison with the gold standard. The second automated model measured the cortical thickness based on a three-dimensional (3D) model rendered from the segmentation results and presented a color visualization of the measurements. The IoU and Hausdorff distance showed high accuracy (0.870 and 0.928 for marrow bone and 0.734 and 1.247 for cortical bone, respectively). A visual comparison of the 3D color maps showed a similar trend to the gold standard. This algorithm for automatic segmentation of the mandibular condyle head and visualization of the measured cortical thickness as a 3D-rendered model with a color map may contribute to the automated quantification of bone thickness changes of the temporomandibular joint complex on CBCT.

Image-based finite-element modeling of the human femur

Fracture is considered a critical clinical endpoint in skeletal pathologies including osteoporosis and bone metastases. However, current clinical guidelines are limited with respect to identifying cases at high risk of fracture, as they do not account for many mechanical determinants that contribute to bone fracture. Improving fracture risk assessment is an important area of research with clear clinical relevance. Patient-specific numerical musculoskeletal models generated from diagnostic images are widely used in biomechanics research and may provide the foundation for clinical tools used to quantify fracture risk. However, prior to clinical translation, in vitro validation of predictions generated from such numerical models is necessary. Despite adopting radically different models, in vitro validation of image-based finite element (FE) models of the proximal femur (predicting strains and failure loads) have shown very similar, encouraging levels of accuracy. The accuracy of such in vitro models has motivated their application to clinical studies of osteoporotic and metastatic fractures. Such models have demonstrated promising but heterogeneous results, which may be explained by the lack of a uniform strategy with respect to FE modeling of the human femur. This review aims to critically discuss the state of the art of image-based femoral FE modeling strategies, highlighting principal features and differences among current approaches. Quantitative results are also reported with respect to the level of accuracy achieved from in vitro evaluations and clinical applications and are used to motivate the adoption of a standardized approach/workflow for image-based FE modeling of the femur.

Cortical bone mapping improves finite element strain prediction accuracy at the proximal femur

Despite evidence of the biomechanical role of cortical bone, current state of the art finite element models of the proximal femur built from clinical CT data lack a subject-specific representation of the bone cortex. Our main research hypothesis is that the subject-specific modelling of cortical bone layer from CT images, through a deconvolution procedure known as Cortical Bone Mapping (CBM, validated for cortical thickness and density estimates) can improve the accuracy of CT-based FE models of the proximal femur, currently limited by partial volume artefacts. Our secondary hypothesis is that a careful choice of cortical-specific density-elasticity relationship may improve model accuracy. We therefore: (i) implemented a procedure to include subject-specific CBM estimates of both cortical thickness and density in CT-based FE models. (ii) defined alternative models that included CBM estimates and featured a cortical-specific or an independently optimised density-elasticity relationship. (iii) tested our hypotheses in terms of elastic strain estimates and failure load and location prediction, by comparing with a published cohort of 14 femurs, where strain and strength in stance and fall loading configuration were experimentally measured, and estimated through reference FE models that did not explicitly model the cortical compartment. Our findings support the main hypothesis: an explicit modelling of the proximal femur cortical bone layer including CBM estimates of cortical bone thickness and density increased the FE strains prediction, mostly by reducing peak errors (average error reduced by 30%, maximum error and 95th percentile of error distribution halved) and especially when focusing on the femoral neck locations (all error metrics at least halved). We instead rejected the secondary hypothesis: changes in cortical density-elasticity relationship could not improve validation performances. From these improved baseline strain estimates, further work is needed to achieve accurate strength predictions, as models incorporating cortical thickness and density produced worse estimates of failure load and equivalent estimates of failure location when compared to reference models. In summary, we recommend including local estimates of cortical thickness and density in FE models to estimate bone strains in physiological conditions, and especially when designing exercise studies to promote bone strength.

Associations of Muscle Size and Density With Proximal Femur Bone in a Community Dwelling Older Population

Background and Purpose: Muscle weakness and bone fragility are both associated with hip fracture. In general, muscle contractions create forces to the bone, and bone strength adapts to mechanical loading through changes in bone architecture and mass. However, the relationship between impairment of muscle and bone function remain unclear. In particular, the associations of muscle with properties of proximal femur cortical and trabecular bone are still not well understood. The aim of this study was to explore the associations of hip/thigh muscle density (CT attenuation value in Hounsfield units) and size with cortical and trabecular bone mineral density (BMD) of the proximal femur. Materials and Methods: Three-dimensional quantitative computed tomography (QCT) imaging of the lumber, hip and mid-thigh was performed in a total of 301 participants (mean age 68.4 ± 6.1 years, 194 women and 107 men) to derive areal BMD (aBMD) and volumetric BMD (vBMD). Handgrip strength (HGS) and the Timed Up and Go (TUG) test were also performed. From the CT images, cross-sectional area (CSA), and density were determined for the gluteus maximus muscle (G.MaxM), trunk muscle at the vertebrae L2 level, and mid-thigh muscle. Multivariate generalized linear models were applied to assess associations. Results: Total hip (TH) aBMD was associated significantly with G.MaxM CSA (men: P = 0.042; women: P < 0.001) and density (men: P = 0.012; women: P = 0.043). In women, 0.035 cm² of mid-thigh CSA (95% CI, 0.014-0.057; P = 0.002) increased per SD increase in TH aBMD, but this significance was not observed in men (P = 0.095). Trunk muscle density and CSA were not associated with proximal femur BMD. The associations of hip/thigh muscle parameters with femoral neck BMD were weaker than those with trochanter and intertrochanter BMD. Furthermore, compared to muscle density, muscle CSA showed better associations with vBMD. G.MaxM CSA was associated with trochanter (TR) Cort. vBMD in men (β, 19.898; 95% CI, 0.924-38.871; P = 0.040) and in women (β, 15.426; 95% CI, 0.893-29.958; P = 0.038). Handgrip strength was only associated with TR aBMD (β, 0.038; 95% CI, 0.006-0.070; P = 0.019) and intertrochanter aBMD (β, 0.049; 95% CI, 0.009-0.090; P = 0.016) in men. Conclusions: We observed positive associations of the gluteus and thigh muscle size with proximal femur volumetric BMD. Specifically, the gluteus maximus muscle CSA was associated with trochanter cortical vBMD in both men and women.

Next-generation imaging of the skeletal system and its blood supply

Bone is organized in a hierarchical 3D architecture. Traditionally, analysis of the skeletal system was based on bone mass assessment by radiographic methods or on the examination of bone structure by 2D histological sections. Advanced imaging technologies and big data analysis now enable the unprecedented examination of bone and provide new insights into its 3D macrostructure and microstructure. These technologies comprise ex vivo and in vivo methods including high-resolution computed tomography (CT), synchrotron-based imaging, X-ray microscopy, ultra-high-field magnetic resonance imaging (MRI), light-sheet fluorescence microscopy, confocal and intravital two-photon imaging. In concert, these techniques have been used to detect and quantify a novel vascular system of trans-cortical vessels in bone. Furthermore, structures such as the lacunar network, which harbours and connects osteocytes, become accessible for 3D imaging and quantification using these methods. Next-generation imaging of the skeletal system and its blood supply are anticipated to contribute to an entirely new understanding of bone tissue composition and function, from macroscale to nanoscale, in health and disease. These insights could provide the basis for early detection and precision-type intervention of bone disorders in the future.

Automated segmentation of cortical and trabecular bone to generate finite element models for femoral bone mechanics

Finite element (FE) models based on quantitative computed tomography (CT) images are better predictors of bone strength than conventional areal bone mineral density measurements. However, FE models require manual segmentation of the femur, which is not clinically applicable. This study developed a method for automated FE analyses from clinical CT images. Clinical in-vivo CT images of 13 elderly female subjects were collected to evaluate the method. Secondly, proximal cadaver femurs were harvested and imaged with clinical CT (N = 17). Of these femurs, 14 were imaged with µCT and three had earlier been tested experimentally in stance-loading, while collecting surface deformations with digital image correlation. Femurs were segmented from clinical CT images using an automated method, based on the segmentation tool Stradwin. The method automatically distinguishes trabecular and cortical bone, corrects partial volume effect and generates input for FE analysis. The manual and automatic segmentations agreed within about one voxel for in-vivo subjects (0.99 ± 0.23 mm) and cadaver femurs (0.21 ± 0.07 mm). The strains from the FE predictions closely matched with the experimentally measured strains (R² = 0.89). The method can automatically generate meshes suitable for FE analysis. The method may bring us one step closer to enable clinical usage of patient-specific FE analyses.

QCT of the femur: Comparison between QCTPro CTXA and MIAF Femur

QCT is commonly employed in research studies and clinical trials to measure BMD at the proximal femur. In this study we compared two analysis software options, QCTPro CTXA and MIAF-Femur, using CT scans of the semi-anthropometric European Proximal Femur Phantom (EPFP) and in vivo data from 130 Chinese elderly men and women aged 60-80 years. Integral (Int), cortical (Cort) and trabecular (Trab) vBMD, volume, and BMC of the neck (FN), trochanter (TR), inter-trochanter (IT), and total hip (TH) VOIs were compared. Accuracy was determined in the 5 mm wide central portion of the femoral neck of the EPFP. Nominal values were: cross-sectional area (CSA) 4.9 cm², cortical thickness (C.Th) 2 mm, CortBMD 723 mg/cm³ and TrabBMD 100 mg/cm³. In MIAF the so-called peeled trabecular VOI was analyzed, which excludes subcortical bone to avoid partial volume artefacts at the endocortical border that artificially increase TrabBMD. For CTXA uncorrected, so called raw cortical values were used for the analysis. QCTPro and MIAF phantom results were: CSA 5.9 cm² versus 5.1 cm²; C.Th 1.68 mm versus 1.92 mm; CortBMD 578 mg/cm³ versus 569 mg/cm³; and TrabBMD 154 mg/cm³ versus 104 mg/cm³. In vivo correlations (R²) of integral and trabecular bone parameters ranged from 0.63 to 0.96. Bland-Altman analysis for TH and FN TrabBMD showed that lower mean values were associated with higher differences, which means that TrabBMD differences between MIAF and CTXA are larger for osteoporotic than for normal patients, which can be largely explained by the inclusion of subcortical BMD in the trabecular VOI analyzed by CTXA in combination with fixed thresholds used to separate cortical from trabecular bone compartments. Correlations between CTXA corrected CortBMD and MIAF were negative, whereas raw data correlated positively with MIAF measurements for all VOIs questioning the validity of the CTXA corrections. The EPFP results demonstrated higher MIAF accuracy of cortical thickness and TrabBMD. Integral and trabecular bone parameters were highly correlated between CTXA and MIAF. Partial volume artefacts at the endocortical border artificially increased trabecular BMD by CTXA, especially for osteoporosis patients. With respect to volumetric cortical measurements with CTXA, the use raw data is recommended, because corrected data cause a negative correlation with MIAF CortBMD.

Bone mineral density averaged over a region of interest on femur is affected by age-related change of bone geometry

Femur expansion occurs during normal aging in both men and women. Average bone mineral density (BMD) over a region of interest (ROI) on the femur may considerably decrease with age even in healthy people, and therefore, it is inaccurate if used to monitor treatment-induced bone change.

INTRODUCTION

Areal bone mineral density (BMD), averaged over a region of interest (ROI) on the femur, is widely used in the diagnosis of osteoporosis, assessment of fracture risk, and monitoring of treatment effectiveness. We studied the effect of age-related change in femur geometry on average BMD.

METHODS

The effect of age-related bone geometric change on averaged BMD was investigated by a cross-sectional study. Total 83 healthy subjects were selected for this study. For each subject, QCT of left femur was scanned using clinical scanner. For each standard volume of interest (VOI), integral/cortical/trabecular bone volume, volumetric BMD (vBMD), and bone mass were measured using QCT Pro; the corresponding areal BMD (aBMD) was projected using CTXA-Hip. Both QCT Pro and CTXA-Hip are commercial software. Correlations between bone volume/density/mass and age were studied.

RESULTS

In the studied population, there was no association between body weight/BMI (body mass index) and age, correlation between normalized femoral neck width and age was 0.24 (p < 0.05). Both aBMD and integral vBMD decreased with age (after adjusted by BMI, for aBMD, r = - 0.21 to - 0.24, p ≤ 0.05 except at trochanter; for vBMD, r = - 0.20 to - 0.31, p < 0.05); cortical vBMD had no significant change; trabecular vBMD decreased at all VOIs except at trochanter (after adjusted by BMI, r = - 0.22 to 0.32, p ≤ 0.05). Integral volume showed slight increase but only significant at the trochanter after adjusted by body size, cortical volume showed insignificant decrease, and trabecular volume considerably increased with age in all VOIs (after adjusted by body size, r = 0.27-0.40, p < 0.05). Integral, cortical, and trabecular mass had no significant change in all VOIs, except that at the trochanter trabecular mass slightly increased with age (r = 0.31, p < 0.05).

CONCLUSIONS

Even though there is no change in bone mass, average BMD may considerably decrease with age due to bone expansion. Comparatively, aBMD is less affected than vBMD.