The potential of multi-slice computed tomography based quantification of the structural anisotropy of vertebral trabecular bone

Micro-computed Tomography-Based Collagen Orientation and Anisotropy Analysis of Rabbit Articular Cartilage

The collagen network is the highly organized backbone of articular cartilage providing tissue tensile stiffness and restricting proteoglycan bleaching out of the tissue. Osteoarthritis (OA) diminishes proper collagen network adaptation. Our aim was to provide quantitative three-dimensional (3D) information of the cartilage collagen network adaptation in early osteoarthritis using high resolution micro-computed tomography (µCT)-imaging. Osteochondral samples from the femoral condyles were collected from healthy (N = 8, both legs) and experimental OA rabbit model with anterior cruciate ligament transection (N = 14, single leg). Samples were processed for cartilage µCT-imaging and histological evaluation with polarized light microscopy (PLM). Structure tensor analysis was used to analyse the collagen fibre orientation and anisotropy of the µCT-images, and PLM was used as a validation for structural changes. Depth-wise comparison of collagen fibre orientation acquired with µCT-imaging and PLM correlated well, but the values obtained with PLM were systematically greater than those measured with µCT-imaging. Structure tensor analysis allowed for 3D quantification of collagen network anisotropy. Finally, µCT-imaging revealed only minor differences between the control and experimental groups.

Directionally Sensitive Fractal Radiomics Compatible With Irregularly Shaped Magnetic Resonance Tumor Regions of Interest: Association With Osteosarcoma Chemoresistance

BACKGROUND

Computational analysis of routinely acquired MRI has potential to improve the tumor chemoresistance prediction and to provide decision support in precision medicine, which may extend patient survival. Most radiomic analytical methods are compatible only with rectangular regions of interest (ROIs) and irregular tumor shape is therefore an important limitation. Furthermore, the currently used analytical methods are not directionally sensitive.

PURPOSE

To implement a tumor analysis that is directionally sensitive and compatible with irregularly shaped ROIs.

STUDY TYPE

Retrospective.

SUBJECTS

A total of 54 patients with histopathologic diagnosis of primary osteosarcoma on tubular long bones and with prechemotherapy MRI.

FIELD STRENGTH/SEQUENCE

A 1.5 T, T2-weighted-short-tau-inversion-recovery-fast-spin-echo.

ASSESSMENT

A model to explore associations with osteosarcoma chemo-responsiveness included MRI data obtained before OsteoSa MAP neoadjuvant cytotoxic chemotherapy. Osteosarcoma morphology was analyzed in the MRI data by calculation of the nondirectional two-dimensional (2D) and directional and nondirectional one-dimensional (1D) Higuchi dimensions (Dh). MAP chemotherapy response was assessed by histopathological necrosis.

STATISTICAL TESTS

The area under the receiver operating characteristic (ROC) curve (AUC) evaluated the association of the calculated features with the actual chemoresponsiveness, using tumor histopathological necrosis (95%) as the endpoint. Least absolute shrinkage and selection operator (LASSO) machine learning and multivariable regression were used for feature selection. Significance was set at <0.05.

RESULTS

The nondirectional 1D Dh reached an AUC of 0.88 in association with the 95% tumor necrosis, while the directional 1D analysis along 180 radial lines significantly improved this association according to the Hanley/McNeil test, reaching an AUC of 0.95. The model defined by variable selection using LASSO reached an AUC of 0.98. The directional analysis showed an optimal predictive range between 90° and 97° and revealed structural osteosarcoma anisotropy manifested by its directionally dependent textural properties.

DATA CONCLUSION

Directionally sensitive radiomics had superior predictive performance in comparison to the standard nondirectional image analysis algorithms with AUCs reaching 0.95 and full compatibility with irregularly shaped ROIs.

EVIDENCE LEVEL

3 TECHNICAL EFFICACY: Stage 1.

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.

Mapping anisotropy improves QCT-based finite element estimation of hip strength in pooled stance and side-fall load configurations

Hip fractures are one of the most severe consequences of osteoporosis. Compared to the clinical standard of DXA-based aBMD at the femoral neck, QCT-based FEA delivers a better surrogate of femoral strength and gains acceptance for the calculation of hip fracture risk when a CT reconstruction is available. Isotropic, homogenised voxel-based, finite element (hvFE) models are widely used to estimate femoral strength in cross-sectional and longitudinal clinical studies. However, fabric anisotropy is a classical feature of the architecture of the proximal femur and the second determinant of the homogenised mechanical properties of trabecular bone. Due to the limited resolution, fabric anisotropy cannot be derived from clinical CT reconstructions. Alternatively, fabric anisotropy can be extracted from HR-pQCT images of cadaveric femora. In this study, fabric anisotropy from HR-pQCT images was mapped onto QCT-based hvFE models of 71 human proximal femora for which both HR-pQCT and QCT images were available. Stiffness and ultimate load computed from anisotropic hvFE models were compared with previous biomechanical tests in both stance and side-fall configurations. The influence of using the femur-specific versus a mean fabric distribution on the hvFE predictions was assessed. Femur-specific and mean fabric enhance the prediction of experimental ultimate force for the pooled, i.e. stance and side-fall, (isotropic: r²=0.81, femur-specific fabric: r²=0.88, mean fabric: r²=0.86,p<0.001) but not for the individual configurations. Fabric anisotropy significantly improves bone strength prediction for the pooled configurations, and mapped fabric provides a comparable prediction to true fabric. The mapping of fabric anisotropy is therefore expected to help generate more accurate QCT-based hvFE models of the proximal femur for personalised or multiple load configurations.

Accounting for spatial variation of trabecular anisotropy with subject-specific finite element modeling moderately improves predictions of local subchondral bone stiffness at the proximal tibia

INTRODUCTION

Previously, a finite element (FE) model of the proximal tibia was developed and validated against experimentally measured local subchondral stiffness. This model indicated modest predictions of stiffness (R²=0.77, normalized root mean squared error (RMSE%)=16.6%). Trabecular bone though was modeled with isotropic material properties despite its orthotropic anisotropy. The objective of this study was to identify the anisotropic FE modeling approach which best predicted (with largest explained variance and least amount of error) local subchondral bone stiffness at the proximal tibia.

METHODS

Local stiffness was measured at the subchondral surface of 13 medial/lateral tibial compartments using in situ macro indentation testing. An FE model of each specimen was generated assuming uniform anisotropy with 14 different combinations of cortical- and tibial-specific density-modulus relationships taken from the literature. Two FE models of each specimen were also generated which accounted for the spatial variation of trabecular bone anisotropy directly from clinical CT images using grey-level structure tensor and Cowin's fabric-elasticity equations. Stiffness was calculated using FE and compared to measured stiffness in terms of R² and RMSE%.

RESULTS

The uniform anisotropic FE model explained 53-74% of the measured stiffness variance, with RMSE% ranging from 12.4 to 245.3%. The models which accounted for spatial variation of trabecular bone anisotropy predicted 76-79% of the variance in stiffness with RMSE% being 11.2-11.5%.

CONCLUSIONS

Of the 16 evaluated finite element models in this study, the combination of Synder and Schneider (for cortical bone) and Cowin's fabric-elasticity equations (for trabecular bone) best predicted local subchondral bone stiffness.

A Predictive Model of Vertebral Trabecular Anisotropy From Ex Vivo Micro-CT

Spine-related disorders are amongst the most frequently encountered problems in clinical medicine. For several applications such as 1) to improve the assessment of the strength of the spine, as well as 2) to optimize the personalization of spinal interventions, image-based biomechanical modeling of the vertebrae is expected to play an important predictive role. However, this requires the construction of computational models that are subject-specific and comprehensive. In particular, they need to incorporate information about the vertebral anisotropic micro-architecture, which plays a central role in the biomechanical function of the vertebrae. In practice, however, accurate personalization of the vertebral trabeculae has proven to be difficult as its imaging in vivo is currently infeasible. Consequently, this paper presents a statistical approach for accurate prediction of the vertebral fabric tensors based on a training sample of ex vivo micro-CT images. To the best of our knowledge, this is the first predictive model proposed and validated for vertebral datasets. The method combines features selection and partial least squares regression in order to derive optimal latent variables for the prediction of the fabric tensors based on the more easily extracted shape and density information. Detailed validation with 20 ex vivo T12 vertebrae demonstrates the accuracy and consistency of the approach for the personalization of trabecular anisotropy.

Micro-computed tomography derived anisotropy detects tumor provoked deviations in bone in an orthotopic osteosarcoma murine model

Radiographic imaging plays a crucial role in the diagnosis of osteosarcoma. Currently, computed-tomography (CT) is used to measure tumor-induced osteolysis as a marker for tumor growth by monitoring the bone fractional volume. As most tumors primarily induce osteolysis, lower bone fractional volume has been found to correlate with tumor aggressiveness. However, osteosarcoma is an exception as it induces osteolysis and produces mineralized osteoid simultaneously. Given that competent bone is highly anisotropic (systematic variance in its architectural order renders its physical properties dependent on direction of load) and that tumor induced osteolysis and osteogenesis are structurally disorganized relative to competent bone, we hypothesized that μCT-derived measures of anisotropy could be used to qualitatively and quantitatively detect osteosarcoma provoked deviations in bone, both osteolysis and osteogenesis, in vivo. We tested this hypothesis in a murine model of osteosarcoma cells orthotopically injected into the tibia. We demonstrate that, in addition to bone fractional volume, μCT-derived measure of anisotropy is a complete and accurate method to monitor osteosarcoma-induced osteolysis. Additionally, we found that unlike bone fractional volume, anisotropy could also detect tumor-induced osteogenesis. These findings suggest that monitoring tumor-induced changes in the structural property isotropy of the invaded bone may represent a novel means of diagnosing primary and metastatic bone tumors.

A novel approach to estimate trabecular bone anisotropy using a database approach

Continuum finite element (FE) models of bones have become a standard pre-clinical tool to estimate bone strength. These models are usually based on clinical CT scans and material properties assigned are chosen as isotropic based only on the density distribution. It has been shown, however, that trabecular bone elastic behavior is best described as orthotropic. Unfortunately, the use of orthotropic models in FE analysis derived from CT scans is hampered by the fact that the measurement of a trabecular orientation (fabric) is not possible from clinical CT images due to the low resolution of such images. In this study, we explore the concept of using a database (DB) of high-resolution bone models to derive the fabric information that is missing in clinical images. The goal of this study was to investigate if models with fabric derived from a relatively small database can already produce more accurate results than isotropic models. A DB of 33 human proximal femurs was generated from micro-CT scans with a nominal isotropic resolution of 82 µm. Continuum FE models were generated from the images using a pre-defined mesh template in combination with an iso-anatomic mesh morphing tool. Each element within the mesh template is at a specific anatomical location. For each element within the cancellous bone, a spherical region around the element centroid with a radius of 2mm was defined. Bone volume fraction and the mean-intercept-length fabric tensor were analyzed for that region. Ten femurs were used as test cases. For each test femur, four different models were generated: (1) an orthotropic model based on micro-CT fabric measurements (gold standard), (2) an orthotropic model based on the fabric derived from the best-matched database model, (3) an isotropic-I model in which the fabric tensor was set to the identity tensor, and (4) a second isotropic-II model with its total bone stiffness fitted to the gold standard. An elastic-plastic damage model was used to simulate failure and post failure behavior during a fall to the side. The results show that all models produce a similar stress distribution. However, compared to the gold standard, both isotropic-I and II models underestimated the stress/damage distributions significantly. We found no significant difference between DB-derived and gold standard models. Compared to the gold standard, the isotropic-I models further underestimated whole bone stiffness by 26.3% and ultimate load by 14.5%, while these differences for the DB-derived orthotropic models were only 4.9% and 3.1% respectively. The results indicate that the concept of using a DB to estimate patient-specific anisotropic material properties can considerably improve the results. We expect that this approach can lead to more accurate results in particular for cases where bone anisotropy plays an important role, such as in osteoporotic patients and around implants.

Measurement of structural anisotropy in femoral trabecular bone using clinical-resolution CT images

Discrepancies in finite-element model predictions of bone strength may be attributed to the simplified modeling of bone as an isotropic structure due to the resolution limitations of clinical-level Computed Tomography (CT) data. The aim of this study is to calculate the preferential orientations of bone (the principal directions) and the extent to which bone is deposited more in one direction compared to another (degree of anisotropy). Using 100 femoral trabecular samples, the principal directions and degree of anisotropy were calculated with a Gradient Structure Tensor (GST) and a Sobel Structure Tensor (SST) using clinical-level CT. The results were compared against those calculated with the gold standard Mean-Intercept-Length (MIL) fabric tensor using micro-CT. There was no significant difference between the GST and SST in the calculation of the main principal direction (median error=28°), and the error was inversely correlated to the degree of transverse isotropy (r=-0.34, p<0.01). The degree of anisotropy measured using the structure tensors was weakly correlated with the MIL-based measurements (r=0.2, p<0.001). Combining the principal directions with the degree of anisotropy resulted in a significant increase in the correlation of the tensor distributions (r=0.79, p<0.001). Both structure tensors were robust against simulated noise, kernel sizes, and bone volume fraction. We recommend the use of the GST because of its computational efficiency and ease of implementation. This methodology has the promise to predict the structural anisotropy of bone in areas with a high degree of anisotropy, and may improve the in vivo characterization of bone.