Intracortical remodelling increases in highly loaded bone after exercise cessation.

Resorption within cortices of long bones removes excess mass and damaged tissue and increases during periods of reduced mechanical loading. Returning to high-intensity exercise may place bones at risk of failure due to increased porosity caused by bone resorption. We used point-projection X-ray microscopy images of bone slices from highly loaded (metacarpal, tibia) and minimally loaded (rib) bones from 12 racehorses, 6 that died during a period of high-intensity exercise and 6 that had a period of intense exercise followed by at least 35 days of rest prior to death, and measured intracortical canal cross-sectional area (Ca.Ar) and number (N.Ca) to infer remodelling activity across sites and exercise groups. Large canals that are the consequence of bone resorption (Ca.Ar >0.04 mm2 ) were 1.4× to 18.7× greater in number and area in the third metacarpal bone from rested than exercised animals (p = 0.005-0.008), but were similar in number and area in ribs from rested and exercised animals (p = 0.575-0.688). An intermediate relationship was present in the tibia, and when large canals and smaller canals that result from partial bony infilling (Ca.Ar >0.002 mm2 ) were considered together. The mechanostat may override targeted remodelling during periods of high mechanical load by enhancing bone formation, reducing resorption and suppressing turnover. Both systems may work synergistically in rest periods to remove excess and damaged tissue.

Classification of racehorse limb radiographs using deep convolutional neural networks.

Purpose: To assess the capability of deep convolutional neural networks to classify anatomical location and projection from a series of 48 standard views of racehorse limbs. Materials and methods: Radiographs (N = 9504) of horse limbs from image sets made for veterinary inspections by 10 independent veterinary clinics were used to train, validate and test (116, 40 and 42 radiographs, respectively) six deep learning architectures available as part of the open source machine learning framework PyTorch. The deep learning architectures with the best top-1 accuracy had the batch size further investigated. Results: Top-1 accuracy of six deep learning architectures ranged from 0.737 to 0.841. Top-1 accuracy of the best deep learning architecture (ResNet-34) ranged from 0.809 to 0.878, depending on batch size. ResNet-34 (batch size = 8) achieved the highest top-1 accuracy (0.878) and the majority (91.8%) of misclassification was due to laterality error. Class activation maps indicated that joint morphology, not side markers or other non-anatomical image regions, drove the model decision. Conclusions: Deep convolutional neural networks can classify equine pre-import radiographs into the 48 standard views including moderate discrimination of laterality, independent of side marker presence.

Closing cones create conical lamellae in secondary osteonal bone.

Lamellae are sheets of mineralized collagen 1-20 µm thick, extending over hundreds of µm in bone tissue, occupying bone's structural hierarchy at a level above collagen fibres and osteocytes, and below osteons and trabeculae. Osteons are tubular arrangements of lamellae surrounding central neurovascular canals. Lamellae in osteons are usually described as concentric cylinders based on their annular appearance in transverse section. In this review, I provide a perspective on current understanding of the relationship between geometry of the bone formation front and the shape of lamellae produced at it, reaching the conclusion that the 'closing cone' bone formation front in secondary osteonal remodelling must necessarily result in cone-shaped lamellae in the mature secondary osteon. Secondary osteons replace primary osteons through a tunnelling process of bone turnover, meaning that conical lamellae may become more common in older and damaged bone which is at greatest risk of fracture. Visualization and measurement of three-dimensional lamellar shape over hundreds of microns is needed to provide data for accurate micromechanical simulations. Treating secondary osteonal lamellae as a 'stack of cones' rather than 'nested cylinders' may have important implications for our appreciation of bone's function as a load-bearing tissue and of its behaviour in fracture.

A Hitchhiker's guide through the bio-image analysis software universe.

Modern research in the life sciences is unthinkable without computational methods for extracting, quantifying and visualising information derived from microscopy imaging data of biological samples. In the past decade, we observed a dramatic increase in available software packages for these purposes. As it is increasingly difficult to keep track of the number of available image analysis platforms, tool collections, components and emerging technologies, we provide a conservative overview of software that we use in daily routine and give insights into emerging new tools. We give guidance on which aspects to consider when choosing the platform that best suits the user's needs, including aspects such as image data type, skills of the team, infrastructure and community at the institute and availability of time and budget.

A New Pipeline to Automatically Segment and Semi-Automatically Measure Bone Length on 3D Models Obtained by Computed Tomography.

The characterization of developmental phenotypes often relies on the accurate linear measurement of structures that are small and require laborious preparation. This is tedious and prone to errors, especially when repeated for the multiple replicates that are required for statistical analysis, or when multiple distinct structures have to be analyzed. To address this issue, we have developed a pipeline for characterization of long-bone length using X-ray microtomography (XMT) scans. The pipeline involves semi-automated algorithms for automatic thresholding and fast interactive isolation and 3D-model generation of the main limb bones, using either the open-source ImageJ plugin BoneJ or the commercial Mimics Innovation Suite package. The tests showed the appropriate combination of scanning conditions and analysis parameters yields fast and comparable length results, highly correlated with the measurements obtained via ex vivo skeletal preparations. Moreover, since XMT is not destructive, the samples can be used afterward for histology or other applications. Our new pipelines will help developmental biologists and evolutionary researchers to achieve fast, reproducible and non-destructive length measurement of bone samples from multiple animal species.

Multithreaded two-pass connected components labelling and particle analysis in ImageJ.

Sequential region labelling, also known as connected components labelling, is a standard image segmentation problem that joins contiguous foreground pixels into blobs. Despite its long development history and widespread use across diverse domains such as bone biology, materials science and geology, connected components labelling can still form a bottleneck in image processing pipelines. Here, I describe a multithreaded implementation of classical two-pass sequential region labelling and introduce an efficient collision resolution step, 'bucket fountain'. Code was validated on test images and against commercial software (Avizo). It was performance tested on images from 2 MB (161 particles) to 6.5 GB (437 508 particles) to determine whether theoretical linear scaling (O(n)) had been achieved, and on 1-40 CPU threads to measure speed improvements due to multithreading. The new implementation achieves linear scaling (b = 0.905-1.052, time ∝ pixelsb ; R 2 = 0.985-0.996), which improves with increasing thread number up to 8-16 threads, suggesting that it is memory bandwidth limited. This new implementation of sequential region labelling reduces the time required from hours to a few tens of seconds for images of several GB, and is limited only by hardware scale. It is available open source and free of charge in BoneJ.

BoneJ2 - refactoring established research software.

Research software is often developed with expedience as a core development objective because experimental results, but not the software, are specified and resourced as a project output. While such code can help find answers to specific research questions, it may lack longevity and flexibility to make it reusable. We reimplemented BoneJ, our software for skeletal biology image analysis, to address design limitations that put it at risk of becoming unusable. We improved the quality of BoneJ code by following contemporary best programming practices. These include separation of concerns, dependency management, thorough testing, continuous integration and deployment, source code management, code reviews, issue and task ticketing, and user and developer documentation. The resulting BoneJ2 represents a generational shift in development technology and integrates with the ImageJ2 plugin ecosystem.

Increased cochlear otic capsule thickness and intracortical canal porosity in the oim mouse model of osteogenesis imperfecta.

Osteogenesis imperfecta (OI or brittle bone disease) is a group of genetic disorders of the connective tissues caused mainly by mutations in the genes encoding collagen type I. Clinical manifestations of OI include skeletal fragility, bone deformities, and severe functional disabilities, such as hearing loss. Progressive hearing loss, usually beginning in childhood, affects approximately 70% of people with OI with more than half of the cases involving the inner ear. There is no cure for OI nor a treatment to ameliorate its corresponding hearing loss, and very little is known about the properties of OI ears. In this study, we investigate the morphology of the otic capsule and the cochlea in the inner ear of the oim mouse model of OI. High-resolution 3D images of 8-week old oim and WT inner ears were acquired using synchrotron microtomography. Volumetric morphometric measurements were conducted for the otic capsule, its intracortical canal network and osteocyte lacunae, and for the cochlear spiral ducts. Our results show that the morphology of the cochlea is preserved in the oim ears at 8 weeks of age but the otic capsule has a greater cortical thickness and altered intracortical bone porosity, with a larger number and volume density of highly branched canals in the oim otic capsule. These results portray a state of compromised bone quality in the otic capsule of the oim mice that may contribute to their hearing loss.

The in vitro behaviour of canine osteoblasts derived from different bone types.

BACKGROUND: Our understanding of the biology of osteoblasts is important as they underpin bone remodelling, fracture healing and processes such as osseointegration. Osteoblasts isolated from human humeral samples display distinctive biological activity in vitro, which relates to the samples' bone types (subchondral (S), trabecular (T), cortical (C)). Our aim was to isolate primary osteoblast cultures from different bone types from the proximal femur of a clinical population of dogs presented for total hip replacement and compare the behaviour of the osteoblasts derived from different bone types, to identify a preferred bone type for isolation. RESULTS: No differences were found for osteoblast doubling time (median for S = 2.9, T = 3.1 and C = 2.71 days, respectively; p = 0.33), final cell number (median for S = 54,849, T = 49,733, C = 61,390 cells/cm2; p = 0.34) or basal tissue non-specific alkaline phosphatase (TNAP) activity (median for S = 0.02, T = 0.02, C = 0.03 U/min/mg protein; p = 0.81) between bone types after 6 days of culture in basal media. There were no differences in mineralizing TNAP activity (S = 0.02, T = 0.02, C = 0.03 U/min/mg protein, p = 0.84) or in mineralized area (S = 0.05, T = 0.04, C = 0.04%, p = 0.92) among cells from different bone types. CONCLUSIONS: There is no significant difference in mean doubling time, basal or mineralizing TNAP activity or mineralized area in osteoblasts derived from subchondral, cortical, or trabecular bone types from the canine femoral head. However, there appears to be a high level of inter-animal variability in the studied parameters, which was independent of age, body mass, and sex. Trabecular isolate osteoblasts have the least variation of the bone types studied, and therefore should be considered a preferred source for primary osteoblast cultures. The work here provides baselines for canine osteoblast function, which has utility for future comparative studies.

Limb bone scaling in hopping macropods and quadrupedal artiodactyls.

Bone adaptation is modulated by the timing, direction, rate and magnitude of mechanical loads. To investigate whether frequent slow, or infrequent fast, gaits could dominate bone adaptation to load, we compared scaling of the limb bones from two mammalian herbivore clades that use radically different high-speed gaits, bipedal hopping (suborder Macropodiformes; kangaroos and kin) and quadrupedal galloping (order Artiodactyla; goats, deer and kin). Forelimb and hindlimb bones were collected from 20 artiodactyl and 15 macropod species (body mass M 1.05-1536 kg) and scanned in computed tomography or X-ray microtomography. Second moment of area (I max) and bone length (l) were measured. Scaling relations (y = axb ) were calculated for l versus M for each bone and for I max versus M and I max versus l for every 5% of length. I max versus M scaling relationships were broadly similar between clades despite the macropod forelimb being nearly unloaded, and the hindlimb highly loaded, during bipedal hopping. I max versus l and l versus M scaling were related to locomotor and behavioural specializations. Low-intensity loads may be sufficient to maintain bone mass across a wide range of species. Occasional high-intensity gaits might not break through the load sensitivity saturation engendered by frequent low-intensity gaits.

Structure Model Index Does Not Measure Rods and Plates in Trabecular Bone.

Structure model index (SMI) is widely used to measure rods and plates in trabecular bone. It exploits the change in surface curvature that occurs as a structure varies from spherical (SMI = 4), to cylindrical (SMI = 3) to planar (SMI = 0). The most important assumption underlying SMI is that the entire bone surface is convex and that the curvature differential is positive at all points on the surface. The intricate connections within the trabecular continuum suggest that a high proportion of the surface could be concave, violating the assumption of convexity and producing regions of negative differential. We implemented SMI in the BoneJ plugin and included the ability to measure the amounts of surface that increased or decreased in area after surface mesh dilation, and the ability to visualize concave and convex regions. We measured SMI and its positive (SMI(+)) and negative (SMI(-)) components, bone volume fraction (BV/TV), the fraction of the surface that is concave (CF), and mean ellipsoid factor (EF) in trabecular bone using 38 X-ray microtomography (XMT) images from a rat ovariectomy model of sex steroid rescue of bone loss, and 169 XMT images from a broad selection of 87 species' femora (mammals, birds, and a crocodile). We simulated bone resorption by eroding an image of elephant trabecule and recording SMI and BV/TV at each erosion step. Up to 70%, and rarely <20%, of the trabecular surface is concave (CF 0.155-0.700). SMI is unavoidably influenced by aberrations induced by SMI(-), which is strongly correlated with BV/TV and CF. The plate-to-rod transition in bone loss is an erroneous observation resulting from the close and artifactual relationship between SMI and BV/TV. SMI cannot discern between the distinctive trabecular geometries typical of mammalian and avian bone, whereas EF clearly detects birds' more plate-like trabecule. EF is free from confounding relationships with BV/TV and CF. SMI results reported in the literature should be treated with suspicion. We propose that EF should be used instead of SMI for measurements of rods and plates in trabecular bone.

The Ellipsoid Factor for Quantification of Rods, Plates, and Intermediate Forms in 3D Geometries.

The ellipsoid factor (EF) is a method for the local determination of the rod- or plate-like nature of porous or spongy continua. EF at a point within a 3D structure is defined as the difference in axis ratios of the greatest ellipsoid that fits inside the structure and that contains the point of interest, and ranges from -1 for strongly oblate (discus-shaped) ellipsoids, to +1 for strongly prolate (javelin-shaped) ellipsoids. For an ellipsoid with axes a ≤ b ≤ c, EF = a/b - b/c. Here, EF is demonstrated in a Java plugin, "Ellipsoid Factor" for ImageJ, distributed in the BoneJ plugin collection. Ellipsoid Factor utilizes an ellipsoid optimization algorithm, which assumes that maximal ellipsoids are centered on the medial axis, then dilates, rotates, and translates slightly each ellipsoid until it cannot increase in volume any further. EF successfully identifies rods, plates, and intermediate structures within trabecular bone, and summarizes the distribution of geometries with an overall EF mean and SD, EF histogram, and Flinn diagram displaying a/b versus b/c. EF is released to the community for testing, use, and improvement.

Four-dimensional in vivo X-ray microscopy with projection-guided gating.

Visualizing fast micrometer scale internal movements of small animals is a key challenge for functional anatomy, physiology and biomechanics. We combine phase contrast tomographic microscopy (down to 3.3 µm voxel size) with retrospective, projection-based gating (in the order of hundreds of microseconds) to improve the spatiotemporal resolution by an order of magnitude over previous studies. We demonstrate our method by visualizing 20 three-dimensional snapshots through the 150 Hz oscillations of the blowfly flight motor.

Iodine vapor staining for atomic number contrast in backscattered electron and X-ray imaging.

Iodine imparts strong contrast to objects imaged with electrons and X-rays due to its high atomic number (53), and is widely used in liquid form as a microscopic stain and clinical contrast agent. We have developed a simple technique which exploits elemental iodine's sublimation-deposition state-change equilibrium to vapor stain specimens with iodine gas. Specimens are enclosed in a gas-tight container along with a small mass of solid I2 . The bottle is left at ambient laboratory conditions while staining proceeds until empirically determined completion (typically days to weeks). We demonstrate the utility of iodine vapor staining by applying it to resin-embedded tissue blocks and whole locusts and imaging them with backscattered electron scanning electron microscopy (BSE SEM) or X-ray microtomography (XMT). Contrast is comparable to that achieved with liquid staining but without the consequent tissue shrinkage, stain pooling, or uneven coverage artefacts associated with immersing the specimen in iodine solutions. Unmineralized tissue histology can be read in BSE SEM images with good discrimination between tissue components. Organs within the locust head are readily distinguished in XMT images with particularly useful contrast in the chitin exoskeleton, muscle and nerves. Here, we have used iodine vapor staining for two imaging modalities in frequent use in our laboratories and on the specimen types with which we work. It is likely to be equally convenient for a wide range of specimens, and for other modalities which generate contrast from electron- and photon-sample interactions, such as transmission electron microscopy and light microscopy.

In vivo time-resolved microtomography reveals the mechanics of the blowfly flight motor.

Dipteran flies are amongst the smallest and most agile of flying animals. Their wings are driven indirectly by large power muscles, which cause cyclical deformations of the thorax that are amplified through the intricate wing hinge. Asymmetric flight manoeuvres are controlled by 13 pairs of steering muscles acting directly on the wing articulations. Collectively the steering muscles account for <3% of total flight muscle mass, raising the question of how they can modulate the vastly greater output of the power muscles during manoeuvres. Here we present the results of a synchrotron-based study performing micrometre-resolution, time-resolved microtomography on the 145 Hz wingbeat of blowflies. These data represent the first four-dimensional visualizations of an organism's internal movements on sub-millisecond and micrometre scales. This technique allows us to visualize and measure the three-dimensional movements of five of the largest steering muscles, and to place these in the context of the deforming thoracic mechanism that the muscles actuate. Our visualizations show that the steering muscles operate through a diverse range of nonlinear mechanisms, revealing several unexpected features that could not have been identified using any other technique. The tendons of some steering muscles buckle on every wingbeat to accommodate high amplitude movements of the wing hinge. Other steering muscles absorb kinetic energy from an oscillating control linkage, which rotates at low wingbeat amplitude but translates at high wingbeat amplitude. Kinetic energy is distributed differently in these two modes of oscillation, which may play a role in asymmetric power management during flight control. Structural flexibility is known to be important to the aerodynamic efficiency of insect wings, and to the function of their indirect power muscles. We show that it is integral also to the operation of the steering muscles, and so to the functional flexibility of the insect flight motor.

Whole-bone scaling of the avian pelvic limb.

Birds form the largest extant group of bipedal animals and occupy a broad range of body masses, from grams to hundreds of kilograms. Additionally, birds occupy distinct niches of locomotor behaviour, from totally flightless strong runners such as the ratites (moa, kiwi, ostrich) to birds that may walk, dabble on water or fly. We apply a whole-bone approach to investigate allometric scaling trends in the pelvic limb bones (femur, tibiotarsus, tarsometatarsus) from extant and recently extinct birds of greatly different size, and compare scaling between birds in four locomotor groups; flightless, burst-flying, dabbling and flying. We also compare scaling of birds' femoral cross-sectional properties to data previously collected from cats. Scaling exponents were not significantly different between the different locomotor style groups, but elevations of the scaling relationships revealed that dabblers (ducks, geese, swans) have particularly short and slender femora compared with other birds of similar body mass. In common with cats, but less pronounced in birds, the proximal and distal extrema of the bones scaled more strongly than the diaphysis, and in larger birds the diaphysis occupied a smaller proportion of bone length than in smaller birds. Cats and birds have similar femoral cross-sectional area (CSA) for the same body mass, yet birds' bone material is located further from the bone's long axis, leading to higher second and polar moments of area and a greater inferred resistance to bending and twisting. The discrepancy in the relationship between outer diameter to CSA may underlie birds' reputation for having 'light' bones.

3D Morphometric and posture study of felid scapulae using statistical shape modelling.

We present a three dimensional (3D) morphometric modelling study of the scapulae of Felidae, with a focus on the correlations between forelimb postures and extracted scapular shape variations. Our shape modelling results indicate that the scapular infraspinous fossa becomes larger and relatively broader along the craniocaudal axis in larger felids. We infer that this enlargement of the scapular fossa may be a size-related specialization for postural support of the shoulder joint.

Osteocyte recruitment declines as the osteon fills in: interacting effects of osteocytic sclerostin and previous hip fracture on the size of cortical canals in the femoral neck.

There is little information on the distribution of osteocytes within the individual cortical osteon, but using direct 3-D imaging in a single subject, Hannah et al. found a gradient with a two-fold higher density of cells adjacent to the cement line compared to near the canal. Since a limiting factor for bone formation might be the availability of osteoblasts due to their recruitment as osteocytes, we studied distributions of osteonal osteocytes in frozen sections of the femoral neck cortex. Osteocytes were stained with an anti-sclerostin antibody and counter-stained with toluidine blue. Adjacent sections were stained for alkaline phosphatase (ALP). Each osteonal osteocyte was categorised as being sclerostin-positive (scl+) or negative (scl-). ImageJ was used to measure the perimeter and area of each osteon and canal, while special purpose routines were used to measure the minimum distances of each osteocyte from the cement line and the canal. Canal area was strongly correlated with osteon area. Osteocytes were most dense close to the cement line; and their areal density within the matrix declined up to three-fold between the cement line and the canal, depending on osteon diameter. Large and small osteons had similar densities of osteocytes close to the cement line, but fractured neck of femur cases had significantly lower densities of osteocytes close to the canal. Higher osteocyte density close to the canal was associated with ALP expression. It is concluded that entombment of osteocytes newly drawn from the osteoblast pool into the mineralising matrix is independent of preceding bone resorption depth. As osteonal infilling proceeds, osteocyte formation declines more rapidly than matrix formation, leading to a progressive reduction in osteocyte density. A shrinking supply of precursor osteoblasts due to previous osteocyte recruitment, apoptosis, or both could produce this effect. In a statistically significant contrast, sclerostin negative osteocytes adjacent to the canal had the expected effect of reducing canal size in controls but this was not seen in hip fracture. This demonstrated the failure of osteonal osteoblasts to sustain bone formation through a complete remodelling cycle in osteoporosis, perhaps due to insufficient osteoblasts remaining capable of mineralized matrix formation. The failure of osteocytic sclerostin suppression to associate with bone formation in these osteons might alternatively be explained by downstream interference with sclerostin's effect on wnt signalling.

Trabecular bone scales allometrically in mammals and birds.

Many bones are supported internally by a latticework of trabeculae. Scaling of whole bone length and diameter has been extensively investigated, but scaling of the trabecular network is not well characterized. We analysed trabecular geometry in the femora of 90 terrestrial mammalian and avian species with body masses ranging from 3 g to 3400 kg. We found that bone volume fraction does not scale substantially with animal size, while trabeculae in larger animals' femora are thicker, further apart and fewer per unit volume than in smaller animals. Finite element modelling indicates that trabecular scaling does not alter the bulk stiffness of trabecular bone, but does alter strain within trabeculae under equal applied loads. Allometry of bone's trabecular tissue may contribute to the skeleton's ability to withstand load, without incurring the physiological or mechanical costs of increasing bone mass.

BoneJ: Free and extensible bone image analysis in ImageJ.

Bone geometry is commonly measured on computed tomographic (CT) and X-ray microtomographic (µCT) images. We obtained hundreds of CT, µCT and synchrotron µCT images of bones from diverse species that needed to be analysed remote from scanning hardware, but found that available software solutions were expensive, inflexible or methodologically opaque. We implemented standard bone measurements in a novel ImageJ plugin, BoneJ, with which we analysed trabecular bone, whole bones and osteocyte lacunae. BoneJ is open source and free for anyone to download, use, modify and distribute.