Growing multiblock structures: a semi-automated approach to block placement for multiblock hexahedral meshing

Finite element (FE) analysis is a cornerstone of orthopaedic biomechanics research. Three-dimensional medical imaging provides sufficient resolution for the subject-specific FE models to be generated from these data-sets. FE model development requires discretisation of a three-dimensional domain, which can be the most time-consuming component of a FE study. Hexahedral meshing tools based on the multiblock method currently rely on the manual placement of building blocks for mesh generation. We hypothesise that angular analysis of the geometric centreline for a three-dimensional surface could be used to automatically generate building block structures for the multiblock hexahedral mesh generation. Our algorithm uses a set of user-defined points and parameters to automatically generate a multiblock structure based on a surface's geometric centreline. This significantly reduces the time required for model development. We have applied this algorithm to 47 bones of varying geometries and successfully generated a FE mesh in all cases. This work represents significant advancement in automatically generating multiblock structures for a wide range of geometries.     

Interactive graph-cut segmentation for fast creation of finite element models from clinical ct data for hip fracture prediction

In this study, we propose interactive graph cut image segmentation for fast creation of femur finite element (FE) models from clinical computed tomography scans for hip fracture prediction. Using a sample of N = 48 bone scans representing normal, osteopenic and osteoporotic subjects, the proximal femur was segmented using manual (gold standard) and graph cut segmentation. Segmentations were subsequently used to generate FE models to calculate overall stiffness and peak force in a sideways fall simulations. Results show that, comparable FE results can be obtained with the graph cut method, with a reduction from 20 to 2–5 min interaction time. Average differences between segmentation methods of 0.22 mm were not significantly correlated with differences in FE derived stiffness (R² = 0.08, p = 0.05) and weakly correlated to differences in FE derived peak force (R² = 0.16, p = 0.01). We further found that changes in automatically assigned boundary conditions as a consequence of small segmentation differences were significantly correlated with FE derived results. The proposed interactive graph cut segmentation software MITK-GEM is freely available online at https://simtk.org/home/mitk-gem.     

Virtual structural analysis of tibial fracture healing from low-dose clinical CT scans

Quantitative assessment of bone fracture healing remains a significant challenge in orthopaedic trauma research. Accordingly, we developed a new technique for assessing bone healing using virtual mechano-structural analysis of computed tomography (CT) scans. CT scans from 19 fractured human tibiae at 12 weeks after surgery were segmented and prepared for finite element analysis (FEA). Boundary conditions were applied to the models to simulate a torsion test that is commonly used to access the structural integrity of long bones in animal models of fracture healing. The output of each model was the virtual torsional rigidity (VTR) of the healing zone, normalized to the torsional rigidity of each patient’s virtually reconstructed tibia. This provided a structural measure to track the percentage of healing each patient had undergone. Callus morphometric measurements were also collected from the CT scans. Results showed that at 12 weeks post-op, more than 75% of patients achieved a normalized VTR (torsional rigidity relative to uninjured bone) of 85% or above. The predicted intact torsional rigidities compared well with published cadaveric data. Across all patients, callus volume and density were weakly and non-significantly correlated with normalized VTR and time to clinical union. Conversely, normalized VTR was significantly correlated with time to union (R² = 0.383, p = 0.005). This suggests that fracture scoring methods based on the visual appearance of callus may not accurately predict mechanical integrity. The image-based structural analysis presented here may be a useful technique for assessment of bone healing in orthopaedic trauma research.     

A semi-automated approach towards generating three-dimensional mesh of the heart using a hybrid MRI/histology database

Both the forward and inverse problems of electrocardiography rely on the precise modelling of the anatomic and electrical properties of the thoracic tissues. This, in turn, requires good knowledge of the electrical anisotropy as well as conductivity inhomogeneity of the heart, lungs and the rest of the thorax. Cardiac electrical anisotropy is related to its microstructure (fibre length, density and orientation). We hereby present detailed three-dimensional (3D) meshes of the thorax and heart, using image data from contiguous 2D magnetic resonance (MR) imaging slices as well as a realistic 3D cardiac fibre orientation model that derives its data from high-resolution ex vivo human heart MR images and from histology specimens of heart tissue. Using specific software, we integrated the 3D thorax and heart meshes in one that addresses the related modelling requirements for the solution of the forward and inverse problems of electrocardiography.     

Gaussian curvature analysis allows for automatic block placement in multi-block hexahedral meshing

Musculoskeletal finite element analysis (FEA) has been essential to research in orthopaedic biomechanics. The generation of a volumetric mesh is often the most challenging step in a FEA. Hexahedral meshing tools that are based on a multi-block approach rely on the manual placement of building blocks for their mesh generation scheme. We hypothesise that Gaussian curvature analysis could be used to automatically develop a building block structure for multi-block hexahedral mesh generation. The Automated Building Block Algorithm incorporates principles from differential geometry, combinatorics, statistical analysis and computer science to automatically generate a building block structure to represent a given surface without prior information. We have applied this algorithm to 29 bones of varying geometries and successfully generated a usable mesh in all cases. This work represents a significant advancement in automating the definition of building blocks.     

A unique feeding strategy of the extinct marine mammal Kolponomos: convergence on sabretooths and sea otters

Mammalian molluscivores feed mainly by shell-crushing or suction-feeding. The extinct marine arctoid, Kolponomos, has been interpreted as an otter-like shell-crusher based on similar dentitions. However, neither the masticatory biomechanics of the shell-crushing adaptation nor the way Kolponomos may have captured hard-shelled prey have been tested. Based on mandibular symphyseal morphology shared by Kolponomos and sabre-toothed carnivores, we hypothesize a sabretooth-like mechanism for Kolponomos prey-capture, whereby the mandible functioned as an anchor. Torque generated from jaw closure and head flexion was used to dislodge prey by prying, with prey then crushed using cheek teeth. We test this hypothesized feeding sequence using phylogenetically informed biomechanical simulations and shape analyses, and find a strongly supported, shared high mandibular stiffness in simulated prey-capture bites and mandibular shape in Kolponomos and the sabre-toothed cat Smilodon. These two distantly related taxa converged on using mandibles to anchor cranial torqueing forces when prying substrate-bound prey in the former and sabre-driving forces during prey-killing in the latter. Simulated prey-crushing bites indicate that Kolponomos and sea otters exhibit alternative structural stiffness-bite efficiency combinations in mandibular biomechanical adaptation for shell-crushing. This unique feeding system of Kolponomos exemplifies a mosaic of form-function convergence relative to other Carnivora.     

Parametric convergence sensitivity and validation of a finite element model of the human lumbar spine

The primary objective of this study was to generate a finite element model of the human lumbar spine (L1–L5), verify mesh convergence for each tissue constituent and perform an extensive validation using both kinematic/kinetic and stress/strain data. Mesh refinement was accomplished via convergence of strain energy density (SED) predictions for each spinal tissue. The converged model was validated based on range of motion, intradiscal pressure, facet force transmission, anterolateral cortical bone strain and anterior longitudinal ligament deformation predictions. Changes in mesh resolution had the biggest impact on SED predictions under axial rotation loading. Nonlinearity of the moment-rotation curves was accurately simulated and the model predictions on the aforementioned parameters were in good agreement with experimental data. The validated and converged model will be utilised to study the effects of degeneration on the lumbar spine biomechanics, as well as to investigate the mechanical underpinning of the contemporary treatment strategies.     

Application of a semi-automatic cartilage segmentation method for biomechanical modeling of the knee joint

Manual segmentation of articular cartilage from knee joint 3D magnetic resonance images (MRI) is a time consuming and laborious task. Thus, automatic methods are needed for faster and reproducible segmentations. In the present study, we developed a semi-automatic segmentation method based on radial intensity profiles to generate 3D geometries of knee joint cartilage which were then used in computational biomechanical models of the knee joint. Six healthy volunteers were imaged with a 3T MRI device and their knee cartilages were segmented both manually and semi-automatically. The values of cartilage thicknesses and volumes produced by these two methods were compared. Furthermore, the influences of possible geometrical differences on cartilage stresses and strains in the knee were evaluated with finite element modeling. The semi-automatic segmentation and 3D geometry construction of one knee joint (menisci, femoral and tibial cartilages) was approximately two times faster than with manual segmentation. Differences in cartilage thicknesses, volumes, contact pressures, stresses, and strains between segmentation methods in femoral and tibial cartilage were mostly insignificant (p > 0.05) and random, i.e. there were no systematic differences between the methods. In conclusion, the devised semi-automatic segmentation method is a quick and accurate way to determine cartilage geometries; it may become a valuable tool for biomechanical modeling applications with large patient groups.     

Semiautomatic extraction of cortical thickness and diaphyseal curvature from CT scans

The understanding of locomotor patterns, activity schemes, and biological variations has been enhanced by the study of the geometrical properties and cortical bone thickness of the long bones measured using CT scan cross‐sections. With the development of scanning procedures, the internal architecture of the long bones can be explored along the entire diaphysis. Recently, several methods that map cortical thickness along the whole femoral diaphysis have been developed. Precise homology is vital for statistical examination of the data; however, the repeatability of these methods is unknown and some do not account for the curvature of the bones. We have designed a semiautomatic workflow that improves the morphometric analysis of cortical thickness, including robust data acquisition with minimal user interaction and considering the bone curvature. The proposed algorithm also performs automatic landmark refinement and rigid registration on the extracted morphometric maps of the cortical thickness. Because our algorithm automatically reslices the diaphysis into 100 cross‐sections along the medial axis and uses an adaptive thresholding method, it is usable on CT scans that contain soft tissues as well as on bones that have not been oriented specifically prior to scanning. Our approach exhibits considerable robustness to error in user‐supplied landmarks, suppresses distortion caused by the curvature of the bones, and calculates the curvature of the medial axis.     

Conditional generative adversarial networks to generate pseudo low monoenergetic CT image from a single-tube voltage CT scanner

PurposeTo generate pseudo low monoenergetic CT images of the abdomen from 120-kVp CT images with cGAN.Materials and MethodsWe retrospectively included 48 patients who underwent contrast-enhanced abdominal CT using dual-energy CT. We reconstructed paired data sets of 120 kVp CT images and virtual low monoenergetic (55-keV) CT images. cGAN was prepared to generate pseudo 55-keV CT images from 120-kVp CT images. The pseudo 55 keV CT images in epoch 10, 50, 100, and 500 were compared to the 55 keV images generated using peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM).ResultsThe PSNRs were 28.0, 28.5, 28.6, and 28.8 at epochs 10, 50, 100, and 500, respectively. The SSIM was approximately constant from epochs 50 to 500.ConclusionPseudo low monoenergetic abdominal CT images were generated from 120-kVp CT images using cGAN, and the images had good quality similar to that of monochromatic images obtained with DECT software.     

A new cortical thickness mapping method with application to an in vivo finite element model

Finite element modelling of musculoskeletal systems, with geometrical structures constructed from computed tomography (CT) scans, is a useful and powerful tool for biomechanical studies. The use of CT scans from living human subjects, however, is still limited. Accurate reconstruction of thin cortical bone structures from CT scans of living human subjects is especially problematic, due to low CT resolution that results from mandatory low radiation doses and/or involuntary movements of the subject. In this study, a new method for mapping cortical thickness is described. Using the method, cortical thickness measurements of a coxal (pelvis) bone obtained from CT scans of a cadaver were mapped to the coxal geometry as obtained through CT scans of a live human subject, resulting in accurate cortical thickness while maintaining geometric fidelity of the live subject. The mapping procedure includes shape-preserving parameterisation, mesh movement and interpolation of thickness using a search algorithm. The methodology is applicable to modelling of other bones where accurate cortical thickness is needed and for which such data exist.     

A modified human head model for the study of impact head injury

A recently published finite element (FE) head model is modified to consider the viscoelasticity of the meninges, the spongy and compact bone in the skull. The cerebrospinal fluid (CSF) is simulated explicitly as a hydrostatic fluid by using a surface-based fluid modelling method, which allows fluid and structure interaction. It is found that the modified model yields smoother pressure responses in a head impact simulation. The baseline model underestimated the peak von Mises stress in the brain by 15% and the peak principal stress in the skull by 33%. The increase in the maximum principal stress in the skull is mainly caused by the updation of the material's viscoelasticity, and the change in the maximum von Mises stress in the brain is mainly caused by the improvement of the CSF simulation. The study shows that the viscoelasticity of the head tissue should be considered, and that the CSF should be modelled as a fluid, when using FE analysis to study head injury due to impact.     

Quantitative comparison of ligament formulation and pre-strain in finite element analysis of the human lumbar spine

Data has been published that quantifies the nonlinear, anisotropic material behaviour and pre-strain behaviour of the anterior longitudinal, supraspinous (SSL), and interspinous ligaments of the human lumbar spine. Additionally, data has been published on localized material properties of the SSL. These results have been incrementally incorporated into a previously validated finite element model of the human lumbar spine. Results suggest that the effects of increased ligament model fidelity on bone strain energy were moderate and the effects on disc pressure were slight, and do not justify a change in modelling strategy for most clinical applications. There were significant effects on the ligament stresses of the ligaments that were directly modified, suggesting that these phenomena should be included in FE models where ligament stresses are the desired metric.     

APPLICATION OF THIN‐PLATE SPLINE TRANSFORMATIONS TO FINITE ELEMENT MODELS,OR, HOW TO TURN A BOG TURTLE INTO A SPOTTED TURTLE TO ANALYZE BOTH

Finite element (FE) models are popular tools that allow biologists to analyze the biomechanical behavior of complex anatomical structures. However, the expense and time required to create models from specimens has prevented comparative studies from involving large numbers of species. A new method is presented for transforming existing FE models using geometric morphometric methods. Homologous landmark coordinates are digitized on the FE model and on a target specimen into which the FE model is being transformed. These coordinates are used to create a thin‐plate spline function and coefficients, which are then applied to every node in the FE model. This function smoothly interpolates the location of points between landmarks, transforming the geometry of the original model to match the target. This new FE model is then used as input in FE analyses. This procedure is demonstrated with turtle shells: a Glyptemys muhlenbergii model is transformed into Clemmys guttata and Actinemys marmorata models. Models are loaded and the resulting stresses are compared. The validity of the models is tested by crushing actual turtle shells in a materials testing machine and comparing those results to predictions from FE models. General guidelines, cautions, and possibilities for this procedure are also presented.     

Pedicle screw-based posterior dynamic stabilisation of the lumbar spine: in vitro cadaver investigation and a finite element study

Pedicle screw-based dynamic constructs either benefit from a dynamic (flexible) interconnecting rod or a dynamic (hinged) screw. Both types of systems have been reported in the literature. However, reports where the dynamic system is composed of two dynamic components, i.e. a dynamic (hinged) screw and a dynamic rod, are sparse. In this study, the biomechanical characteristics of a novel pedicle screw-based dynamic stabilisation system were investigated and compared with equivalent rigid and semi-rigid systems using in vitro testing and finite element modelling analysis. All stabilisation systems restored stability after decompression. A significant decrease in the range of motion was observed for the rigid system in all loadings. In the semi-rigid construct the range of motion was significantly less than the intact in extension, lateral bending and axial rotation loadings. There were no significant differences in motion between the intact spine and the spine treated with the dynamic system (P>0.05). The peak stress in screws was decreased when the stabilisation construct was equipped with dynamic rod and/or dynamic screws.     

Factors affecting subject-specific finite element models of implant-fitted rat bone specimens: critical analysis of a technical protocol

The authors propose a protocol to derive finite element (FE) models from micro computer tomography scans of implanted rat bone. A semi-automatic procedure allows segmenting the images using specimen-specific bone mineral density (BMD) thresholds. An open-source FE model generator processes the segmented images to a quality tetrahedral mesh. The material properties assigned to each element are integrated from the BMD field. Piecewise, threshold-dependent density–elasticity relationships are implemented to limit the effects of metal artefacts. A detailed sensitivity study highlights the coherence of the generated models and quantifies the influence of the modelling parameters on the results. Two applications of the protocol are proposed. The stiffness of bare and implanted rat tibiae specimens is predicted by simulating three-point bending and inter-implant displacement, respectively. Results are compared with experimental tests. The mean value and the variability between the specimens are well captured in both tests.