Cortical Bone Mapping: Measurement and Statistical Analysis of Localised Skeletal Changes

Musculoskeletal characteristics in older adults with overweight or obesity: INVEST in Bone Health trial baseline analysis

OBJECTIVE

The objective of this study was to examine associations of computed tomography (CT)-derived musculoskeletal measures with demographics and traditional musculoskeletal characteristics.

METHODS

The Incorporating Nutrition, Vests, Education, and Strength Training (INVEST) in Bone Health trial (NCT04076618) acquired a battery of musculoskeletal measures in 150 older-aged adults living with overweight or obesity. At baseline, CT (i.e., volumetric bone mineral density, cortical thickness, muscle radiomics, and muscle/intermuscular adipose tissue [IMAT] area and density), dual-energy x-ray absorptiometry (DXA; i.e., areal bone mineral density, total body fat mass, appendicular lean mass, and lean body mass), and strength assessments (i.e., grip and knee extensor strength) were collected, along with demographic and clinical characteristics. Analyses employed linear regression and mixed-effects models along with factor analysis for dimensionality reduction of the radiomics data.

RESULTS

Participants were older-aged (mean [SD] age: 66 [5] years), mostly female (75%), and were living with overweight or obesity (mean [SD] BMI: 33.6 [3.3] kg/m2). Age was not significantly associated with most CT-derived bone, IMAT, or muscle measures. BMI was significantly associated with DXA and CT-derived muscle and IMAT measures, which were higher in male than female individuals (all p < 0.01). For the midthigh, muscle size was significantly related to grip and knee extensor strength (both p < 0.01).

CONCLUSIONS

Machine learning-derived CT metrics correlated strongly with DXA and muscle strength, with higher BMI linked to greater IMAT and poorer muscle quality.

HR-pQCT measurements of changes in periarticular bone density and microarchitecture one year after acute knee injury and after reconstructive surgery

ACL injuries commonly lead to post-traumatic osteoarthritis (PTOA), but the underlying mechanism is not well-understood. One theorized mechanism is pathological bone remodelling following an ACL tear, for which high-resolution peripheral quantitative computed tomography (HR-pQCT) is uniquely positioned to investigate in vivo in humans. In this study, we longitudinally investigate the one-year changes in periarticular bone density and microarchitecture in the human knee following an ACL tear and reconstructive surgery using data sampled from an on-going observational cohort study. We reduce the number of individual microarchitectural parameters using factor analysis and model one-year changes with mixed-effects models, adjusting for the effects of age, sex, meniscus status, and the baseline microarchitectural state. We find significant evidence of persistent bone density losses one year after both injury and surgery. We also observe significant increases in trabecular separation post-injury, indicating significant structural degradation, and significant increases in subchondral bone plate density post-surgery, a sign of early stiffening. Finally, we observe minimal significant contrasts for the effects of age, sex, and meniscus status, while we observe that the state of the microarchitecture at baseline has significant and varied effects on the subsequent changes, suggesting that the influence of PTOA risk factors on post-injury and post-surgery bone changes may be mediated through the state of the periarticular microarchitecture at injury and/or at surgery. In summary, we found that degradation of periarticular bone microarchitecture was observed post-injury, densification of the subchondral bone plate was observed post-surgery, and the state of the bone microarchitecture at baseline may mediate the influence of PTOA risk factors on post-injury microarchitectural adaptations.

A data-driven framework for developing a unified density-modulus relationship for the human lumbar vertebral body

Despite the broad agreement that bone stiffness is heavily dependent on the underlying bone density, there is no consensus on a unified relationship that applies to both cancellous and cortical compartments. Bone from the two compartments is generally assessed separately, and few mechanical test data are available for samples from the transitional regions between them. In this study, we present a data-driven framework integrating experimental testing and numerical modeling of the human lumbar vertebra through an energy balance criterion, to develop a unified density-modulus relationship across the entire vertebral body, without the necessity of differentiation between trabecular and cortical regions. A dataset of 25 spinal segments harvested from fresh-frozen human spines consisting of L1 vertebrae with adjacent intervertebral disks and neighboring T12 and L2 endplates was examined through a systematic process. Each specimen was subjected to axial compression using a custom-designed radiolucent device, and the deformation at multiple points during the ramp was quantified using digital volume correlation applied to the time-lapse series of microcomputed tomography images acquired during loading. A finite element model of each specimen was constructed from quantitative computed tomography images, with the experimental displacement fields imposed to replicate the observed deformation. The optimal density-modulus relationship, both in exponential and polynomial forms, was then determined by using data-driven techniques to match the numerical strain energy with the experimental external work. The resulting relationships effectively recovered bone tissue modulus at the microscale. Subsequently, the unified relationships were applied to investigate the vertebral structure-property correlations at the macroscale: as expected, compressive stiffness exhibited a moderate correlation with bone mineral density, whereas bending stiffness was revealed to correlate strongly with bone mineral content. These findings support the accuracy of the developed density-modulus relationships for the vertebral body and indicate the potential of the proposed framework to extend to other properties of interest such as vertebral strength and toughness.

Generation of Customized Bone Implants from CT Scans Using FEA and AM

Additive manufacturing (AM) allows the creation of customized designs for various medical devices, such as implants, casts, and splints. Amongst other AM technologies, fused filament fabrication (FFF) facilitates the production of intricate geometries that are often unattainable through conventional methods like subtractive manufacturing. This study aimed to develop a methodology for substituting a pathological talus bone with a personalized one created using additive manufacturing. The process involved generating a numerical parametric solid model of the specific anatomical region using computed tomography (CT) scans of the corresponding healthy organ from the patient. The healthy talus served as a mirrored template to replace the defective one. Structural simulation of the model through finite element analysis (FEA) helped compare and select different materials to identify the most suitable one for the replacement bone. The implant was then produced using FFF technology. The developed procedure yielded commendable results. The models maintained high geometric accuracy, while significantly reducing the computational time. PEEK emerged as the optimal material for bone replacement among the considered options and several specimens of talus were successfully printed.

Preventing stress singularities in peri-implant bone - a finite element analysis using a graded bone model

In finite element analysis bone is often treated as two-layered material that has a discontinuity between the cortical and cancellous bone, which leads to a singularity and incorrect stresses. The goal of this study was to eliminate this singularity and to create a more realistic representation of bone which also considers the transition zone between cortical and cancellous bone as observed in natural bone. This was achieved by modelling bone as a graded material and inserting node-specific values for Young's modulus in the finite element simulation, whereas the transition zone thickness was derived from a CT scan. The modelling was performed semi-automatically, and the maximum principal stresses of the new approach were compared to those of a conventional approach. The new approach was found to effectively avoid singularities and provides more accurate predictions of stress in areas of the bone transition zone. As the approach is automatable and causes rather small overhead it is recommended for use in future work, when the problem at hand requires evaluating stresses close to the former singularity.

Mapping the Spatial Evolution of Proximal Femur Osteoporosis: A Retrospective Cross-Sectional Study Based on CT Scans

Purpose

The purpose of this study was to quantify the modifications occurring in osteoporosis at the level of the human proximal femur throughout the trabecular structure, along with the identification of certain anatomic regions preferentially affected by osteoporosis. Another goal was to map the evolution of the radiodensity of the trabecular bone as osteoporosis progresses to an advanced stage.

Methods

The study included CT scans (right femur) from 51 patients, out of which 40 had various degrees of osteoporosis, but no other local pathology. Ten regions of interest in two orthogonal slices have been identified and the differences in radiodensity as well as their evolution have been statistically analyzed in terms of relative and absolute changes.

Results

A detailed spatial map showing the evolution of osteoporosis was obtained. As osteoporosis evolved, the relative decrease in radiodensity was inversely correlated to the radiodensity of the healthy bone. In particular, the region covering the Ward triangle decreased the most, by an average 61-62% in osteopenia and 101-106% in advanced osteoporosis, while the principal compressive group was affected the least, showing a decrease by an average 14-15% in osteopenia and 29-32% in advanced osteoporosis. The absolute decrease in radiodensity was not correlated to the radiodensity of the healthy bone and was shifted to the inferior-posterior edge of the femur. Inside the femoral head, the upper region was affected the most in absolute terms, while the greater trochanter was less affected than the femoral neck. The maximum metaphyseal cortical bone density was unaffected by the progression of osteoporosis.

Conclusion

Significant differences were noticed in terms of the absolute and relative osteoporotic changes in radiodensity related to different anatomical regions of the human femoral bone. These differences become more pronounced as the disease progresses.

Cortical and Trabecular Bone Deficit in Middle-Aged Men Living with HIV

A significant increase in the risk of hip fracture was observed in middle-aged men living with human immunodeficiency virus (MLWH), almost a decade earlier than those without infection. Data regarding cortical and trabecular bone deficit of hip, an important determinant of bone strength, in MLWH are limited. Quantitative CT was performed in consecutive MLWH aged ≥30 years between November 2017 and October 2018 at Severance Hospital, Seoul, Korea. Volumetric bone mineral density (vBMD) and cortical bone mapping parameters of hip (cortical thickness [CTh], cortical bone vBMD [CBMD], cortical mass surface density [CMSD], endocortical trabecular density [ECTD]) were compared to age-matched and body mass index (BMI)-matched controls (1:2) using a community-based healthy adults cohort. Among 83 MLWH and 166 controls (mean age: 47.2 years; BMI: 23.6 kg/m2 ), MLWH had lower total hip vBMD (280 ± 41 versus 296 ± 41 mg/cm3 ), CMSD (155 versus 160 mg/cm2 ), and ECTD (158 versus 175 mg/cm3 ) than controls that remained robust after adjustment for covariates (adjusted β: total hip vBMD, -18.8; CMSD, -7.3; ECTD, -18.0; p < 0.05 for all). Cortical bone mapping revealed localized deficit of CTh, CBMD, and CMSD in the anterolateral trochanteric region and femoral neck in MLWH compared to controls, with a more extensive ECTD deficit. In MLWH, lower CD4 T-cell count (/100 cells/mm3 decrement) and protease inhibitor (PI)-based regimen (versus non-PI regimen) at the time of antiretroviral treatment initiation were associated with lower total hip vBMD (adjusted β -7.5 for lower CD4 count; -28.3 for PI-based regimen) and CMSD (adjusted β -2.6 for lower CD4 count; -12.7 for PI-based regimen; p < 0.05 for all) after adjustment for covariates including age, BMI, smoking, alcohol use, hepatitis C virus co-infection, tenofovir exposure, and CT scanner types. MLWH had lower hip bone density with cortical and trabecular bone deficit compared to community-dwelling controls. © 2023 American Society for Bone and Mineral Research (ASBMR).

Systematic measuring cortical thickness in tibiae for bio-mechanical analysis

BACKGROUND AND OBJECTIVE

Measuring the thickness of cortical bone tissue helps diagnose bone diseases or monitor the progress of different treatments. This type of measurement can be performed visually from CAT images by a radiologist or by semi-automatic algorithms from Hounsfield values. This article proposes a mechanism capable of measuring thickness over the entire bone surface, aligning and orienting all the images in the same direction to have comparable references and reduce human intervention to a minimum. The objective is to batch process large numbers of patients' CAT images obtaining thicknesses profiles of their cortical tissue to be used in many applications.

METHODS

Classical morphological and Deep Learning segmentation is used to extract the area of interest, filtering and interpolation to clean the bones and contour detection and Signed Distance Functions to measure the cortical Thickness. The alignment of the set of bones is achieved by detecting their longitudinal direction, and the orientation is performed by computing their principal component of the center of mass slice.

RESULTS

The method processed in an unattended manner 67% of the patients in the first run and 100% in the second run. The difference in the thickness values between the values provided by the algorithm and the measures done by a radiologist was, on average, 0.25 millimetres with a standard deviation of 0.2.

CONCLUSION

Measuring the cortical thickness of a bone would allow us to prepare accurate traumatological surgeries or study their structural properties. Obtaining thickness profiles of an extensive set of patients opens the way for numerous studies to be carried out to find patterns between bone thickness and the patients' medical, social or demographic variables.

Evaluation of Serum Albumin-Coated Bone Allograft for Bone Regeneration: A Seven-Year Follow-Up Study of 26 Cases

We have previously reported that serum albumin-coated bone allograft (BoneAlbumin, BA) is an effective bone substitute. It improves bone regeneration at the patellar and tibial donor sites six months after harvesting bone-patellar tendon-bone (BPTB) autografts for primary anterior cruciate ligament reconstruction (ACLR). In the present study, we examined these donor sites seven years after implantation. The study group (N = 10) received BA-enhanced autologous cancellous bone at the tibial and BA alone at the patellar site. The control group (N = 16) received autologous cancellous bone at the tibial and blood clot at the patellar site. We evaluated subcortical density, cortical thickness, and bone defect volume via CT scans. At the patellar site, subcortical density was significantly higher in the BA group at both time points. There was no significant difference in cortical thickness between the two groups at either donor site. The control group's bone defect significantly improved and reached the BA group's values at both sites by year seven. Meanwhile, the bone defects in the BA group did not change significantly and were comparable to the six-month measurements. No complications were observed. There are two limitations in this study: The number of patients recruited is small, and the randomization of the patients could have improved the quality of the study as the control group patients were older compared to the study group patients. Our 7-year results seem to demonstrate that BA is a safe and effective bone substitute that supports faster regeneration of donor sites and results in good-quality bone tissue at the time of ACLR with BPTB autografts. However, studies with a larger number of patients are required to definitively confirm the preliminary results of our study.

Intrinsic Cortical Property Analysis of the Medial Column of Proximal Humerus

OBJECTIVE

Adequate mechanical support of the medial column is paramount to maintain fracture reduction in locking plating of proximal humerus fractures. However, intrinsic cortical properties of the medial column are rarely discussed. The purpose of the study is to describe regional variation of cortex in the medial column.

METHODS

A total of 147 healthy participants were eligible for enrollment between December 2016 and December 2018. Subjects were divided into three groups: group A (20-39 years), group B (40-59 years), and group C (>60 years). For each individual, a color 3D thickness map for proximal humerus was created by cortical bone mapping (CBM) technique after bilateral shoulders were imaged by computed tomography. Measurement Indices including the cortical thickness (CTh), cortical mass surface density (CM) and the endocortical trabecular density (ECTD) were determined, after six regions of interest (ROI) were defined in metaphyseal region. Regional parameter variations were analyzed by one-way ANOVA.

RESULTS

The CTh, CM and ECTD values were approximately equivalent between genders in the proximal part of the medial column across all ages (P > 0.05).The greatest difference between sexes was found in CTh and CM values of middle and distal part (P < 0.05). The CTh and CM within medial column were negatively associated with age (P < 0.05). The proximal cortical bone of the medial column was thicker and more dense, compared to the lateral column (P < 0.05). Significant regional variation was found in all measured parameters in group A, but not in groups B and C.

CONCLUSION

Our finding proved that regional differences in the distribution of cortical bone in the medial column The attenuation of cortical bone heterogeneity in the medial column was found after the age of 40 years.

The Influence of Environmental Exposure to Heavy Metals on the Occurrence of Selected Elements in the Maxillary Bone

The elemental composition of the body's calcified tissues may reflect the environmental exposure of the population to heavy metals. The aim of the study was to assess whether the elemental composition of the maxillary bone from individuals belonging to a given population reflects the environmental exposure of this population to lead and cadmium. The research material consisted of cortical bone from the anterolateral walls of the maxilla collected from 126 patients during Caldwell-Luc maxillary sinus surgery on residents of two cities differing in terms of the lead and cadmium pollution of the natural environment. The content levels of lead, cadmium, iron, manganese, chromium, copper, and iron were determined by using atomic absorption spectrophotometry. The content levels of lead and cadmium in the samples of the maxillary bones of residents of Bielsko-Biala were 3.26 ± 2.42 µg/g and 0.74 ± 0.38 µg/g, respectively, whereas in the samples from the residents of Katowice, they were 7.66 ± 2.79 µg/g and 1.12 ± 0.08 µg/g, respectively. It was found that the lead and cadmium levels in the maxillary bone corresponded to the environmental exposure to these heavy metals in the place of residence, which was proven here via the example of the residents of two cities with different concentrations of these heavy metals in the air over long time periods. Additionally, higher content levels of essential metals such as manganese, chromium, copper, and iron are characteristic of the maxillary bone samples of residents of the area that is more polluted with heavy metals.

Prediction of femoral strength of elderly men based on quantitative computed tomography images using machine learning

Hip fracture is the most common complication of osteoporosis, and its major contributor is compromised femoral strength. This study aimed to develop practical machine learning models based on clinical quantitative computed tomography (QCT) images for predicting proximal femoral strength. Eighty subjects with entire QCT data of the right hip region were randomly selected from the full MrOS cohorts, and their proximal femoral strengths were calculated by QCT-based finite element analysis (QCT/FEA). A total of 50 parameters of each femur were extracted from QCT images as the candidate predictors of femoral strength, including grayscale distribution, regional cortical bone mapping (CBM) measurements, and geometric parameters. These parameters were simplified by using feature selection and dimensionality reduction. Support vector regression (SVR) was used as the machine learning algorithm to develop the prediction models, and the performance of each SVR model was quantified by the mean squared error (MSE), the coefficient of determination (R² ), the mean bias, and the SD of bias. For feature selection, the best prediction performance of SVR models was achieved by integrating the grayscale value of 30% percentile and specific regional CBM measurements (MSE ≤ 0.016, R² ≥ 0.93); and for dimensionality reduction, the best prediction performance of SVR models was achieved by extracting principal components with eigenvalues greater than 1.0 (MSE ≤ 0.014, R² ≥ 0.93). The femoral strengths predicted from the well-trained SVR models were in good agreement with those derived from QCT/FEA. This study provided effective machine learning models for femoral strength prediction, and they may have great potential in clinical bone health assessments.

Subchondral bone changes after joint distraction treatment for end stage knee osteoarthritis

OBJECTIVE

Increased subchondral cortical bone plate thickness and trabecular bone density are characteristic of knee osteoarthritis (OA). Knee joint distraction (KJD) is a joint-preserving knee OA treatment where the joint is temporarily unloaded. It has previously shown clinical improvement and cartilage regeneration, indicating reversal of OA-related changes. The purpose of this research was to explore 3D subchondral bone changes after KJD treatment using CT imaging.

DESIGN

Twenty patients were treated with KJD and included to undergo knee CT imaging before, one, and two years after treatment. Tibia and femur segmentation and registration to canonical surfaces were performed semi-automatically. Cortical bone thickness and trabecular bone density were determined using an automated algorithm. Statistical parametric mapping (SPM) with two-tailed F-tests was used to analyze whole-joint changes.

RESULTS

Data was available of 16 patients. Subchondral cortical bone plate thickness and trabecular bone density were higher in the weight-bearing region of the most affected compartment (MAC; mostly medial). Especially the MAC showed a decrease in thickness and density in the first year after treatment, which was sustained towards the second year.

CONCLUSIONS

KJD treatment results in bone changes that include thinning of the subchondral cortical bone plate and decrease of subchondral trabecular bone density in the first two years after treatment, potentially indicating a partial normalization of subchondral bone.

Multiparametric 3-D analysis of bone and joint space width at the knee from weight bearing computed tomography

OBJECTIVE

Computed tomography (CT) can deliver multiple parameters relevant to osteoarthritis. In this study we demonstrate that a 3-D multiparametric approach at the weight bearing knee with cone beam CT is feasible, can include multiple parameters from across the joint space, and can reveal stronger relationships with disease status in combination.

DESIGN

33 participants with knee weight bearing CT (WBCT) were analysed with joint space mapping and cortical bone mapping to deliver joint space width (JSW), subchondral bone plate thickness, endocortical thickness, and trabecular attenuation at both sides of the joint. All data were co-localised to the same canonical surface. Statistical parametric mapping (SPM) was applied in uni- and multivariate models to demonstrate significant dependence of parameters on Kellgren & Lawrence grade (KLG). Correlation between JSW and bony parameters and 2-week test-retest repeatability were also calculated.

RESULTS

SPM revealed that the central-to-posterior medial tibiofemoral joint space was significantly narrowed by up to 0.5 mm with significantly higher tibial trabecular attenuation up to 50 units for each increment in KLG as single features, and in a wider distribution when combined (p<0.05). These were also more strongly correlated with worsening KLG grade category. Test-retest repeatability was subvoxel (0.37 mm) for nearly all thickness parameters.

CONCLUSIONS

3-D JSW and tibial trabecular attenuation are repeatable and significantly dependent on radiographic disease severity at the weight bearing knee joint not just alone, but more strongly in combination. A quantitative multiparametric approach with WBCT may have potential for more sensitive investigation of disease progression in osteoarthritis.

Romosozumab Enhances Vertebral Bone Structure in Women With Low Bone Density

Romosozumab monoclonal antibody treatment works by binding sclerostin and causing rapid stimulation of bone formation while decreasing bone resorption. The location and local magnitude of vertebral bone accrual by romosozumab and how it compares to teriparatide remains to be investigated. Here we analyzed the data from a study collecting lumbar computed tomography (CT) spine scans at enrollment and 12 months post-treatment with romosozumab (210 mg sc monthly, n = 17), open-label daily teriparatide (20 μg sc, n = 19), or placebo (sc monthly, n = 20). For each of the 56 women, cortical thickness (Ct.Th), endocortical thickness (Ec.Th), cortical bone mineral density (Ct.bone mineral density (BMD)), cancellous BMD (Cn.BMD), and cortical mass surface density (CMSD) were measured across the first lumbar vertebral surface. In addition, color maps of the changes in the lumbar vertebrae structure were statistically analyzed and then visualized on the bone surface. At 12 months, romosozumab improved all parameters significantly over placebo and resulted in a mean vertebral Ct.Th increase of 10.3% versus 4.3% for teriparatide, an Ec.Th increase of 137.6% versus 47.5% for teriparatide, a Ct.BMD increase of 2.1% versus a -0.1% decrease for teriparatide, and a CMSD increase of 12.4% versus 3.8% for teriparatide. For all these measurements, the differences between romosozumab and teriparatide were statistically significant (p < 0.05). There was no significant difference between the romosozumab-associated Cn.BMD gains of 22.2% versus 18.1% for teriparatide, but both were significantly greater compared with the change in the placebo group (-4.6%, p < 0.05). Cortical maps showed the topographical locations of the increase in bone in fracture-prone areas of the vertebral shell, walls, and endplates. This study confirms widespread vertebral bone accrual with romosozumab or teriparatide treatment and provides new insights into how the rapid prevention of vertebral fractures is achieved in women with osteoporosis using these anabolic agents. © 2021 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Protective effect of bisphosphonate on the cortical bone at key locations of the femur in aromatase inhibitor-associated bone loss: A three-dimensional cortical bone mapping study

•

Aromatase inhibitor use was associated with cortical bone loss in the hip.

•

Bisphosphonate protected hip cortical bone against aromatase inhibitor use.

•

The effect was prominent at the superior femoral neck and greater trochanter.

Aromatase inhibitor treatment in breast cancer is associated with accelerated bone loss and an increased risk of fracture. Bisphosphonates (BPs) are the mainstay treatment of aromatase inhibitor-associated bone loss (AIBL), which might improve femoral bone at key locations prone to fracture. To test this hypothesis, we performed three-dimensional cortical bone mapping based on quantitative computed tomography (QCT) scans in postmenopausal women with early breast cancer who were receiving aromatase inhibitors. Data of subjects who had both baseline and at least one follow-up QCT at Severance Hospital (South Korea) between 2005 and 2015 were analyzed (BP users, n = 93; BP non-users, n = 203). After exclusion of BP users with low medication persistence (proportion of days covered: <50%), BP users and non-users were 1:1 matched (n = 54 for each group) in terms of age, lumbar spine volumetric bone mineral density (LSvBMD), femoral neck areal BMD (FNaBMD), and total hip areal BMD (THaBMD). During a median follow-up of 2.1 years, BP use attenuated bone loss in LSvBMD (+7.2% vs. −3.8%, p < 0.001), FNaBMD (+1.3% vs. −2.7%, p < 0.001), and THaBMD (-0.3% vs. −2.5%, p = 0.024). BP had a protective effect on cortical parameters of femoral bone: estimated cortical thickness (CTh) (+3.3% vs. + 0.1%, p = 0.007) and cortical mass surface density (CMSD, cortical mass per unit surface area was calculated by multiplying cortical BMD with CTh) (+3.4% vs. −0.3%, p < 0.001). CMSD increased by up to 15% at key locations such as the superior part of the femoral neck and greater trochanter. BP prevented the thinning of average CTh of the femoral neck (-1.4% vs. −6.1%, p < 0.001), particularly at the superior anterior quadrant of femoral neck (absolute difference: +12.8% point vs. non-users). Compared to BP non-users, BP users had improved cross-sectional moment of inertia (+4.4% vs. −0.7%, p = 0.001) and less increase in buckling ratio (+1.3% vs. + 7.5%, p < 0.001). In summary, BP use prevented cortical bone deficits observed in AIBL at key locations of the proximal femur.

Augmenting Osteoporosis Imaging with Machine Learning

PURPOSE OF REVIEW

In this paper, we discuss how recent advancements in image processing and machine learning (ML) are shaping a new and exciting era for the osteoporosis imaging field. With this paper, we want to give the reader a basic exposure to the ML concepts that are necessary to build effective solutions for image processing and interpretation, while presenting an overview of the state of the art in the application of machine learning techniques for the assessment of bone structure, osteoporosis diagnosis, fracture detection, and risk prediction.

RECENT FINDINGS

ML effort in the osteoporosis imaging field is largely characterized by "low-cost" bone quality estimation and osteoporosis diagnosis, fracture detection, and risk prediction, but also automatized and standardized large-scale data analysis and data-driven imaging biomarker discovery. Our effort is not intended to be a systematic review, but an opportunity to review key studies in the recent osteoporosis imaging research landscape with the ultimate goal of discussing specific design choices, giving the reader pointers to possible solutions of regression, segmentation, and classification tasks as well as discussing common mistakes.

The issues and complexities of establishing methodologies to differentiate between vertical and horizontal impact mechanisms in the analysis of skeletal trauma: An introductory femoral test

Understanding skeletal trauma characteristics is fundamental for the examination and interpretation of blunt force trauma (BFT). BFT is the most complex type of trauma to interpret based on the analysis of skeletal fractures alone, with comminuted fractures presenting additional complications to assess and interpret. Considerable variation exists within each type of BFT injury dependent on direction, magnitude of force, plus a myriad of biological/environmental factors. Given the complex processes governing the nature of BFT skeletal injuries determining whether differences between impact mechanisms and skeletal trauma can be quantified requires investigation.

AIM

this study aims to determine the feasibility of quantifying outcomes between two separate loading conditions by using a formula created from transformed variables recorded from specific trauma cases involving BFT to the femur.

METHODOLOGY

Displacement, comminution, and femoral midshaft area data were recorded from full body postmortem computed tomography scans of 103 individuals (males, mean age 42.5, and females, mean age 48.9) where cause of death was the result of rapid horizontal deceleration impact events (pedestrian motor vehicular accidents, n = 59) and vertical (>3-metre falls, n = 44). These measurements were standardised and transformed into a continuous variable. Independent t-tests, binary logistic regression and K Nearest- Neighbours (KNN) were used to analyse the data.

RESULTS

The standardised values showed mean group differences between falls (9.62) and pedestrian motor vehicular impacts (pedestrian MVAs) (9.53), however, these results were not statistically significant. The results indicate that similarities in variance between types of trauma outcomes and impact mechanisms demonstrate low equivalency (samples have limited differences), and the overall limitations in relying on using single elements to explain complex skeletal trauma outcomes.

3D DXA Hip Differences in Patients with Acromegaly or Adult Growth Hormone Deficiency

The skeleton is regulated by and responds to pituitary hormones, especially when the circulating levels are perturbed in disease. This study aims to analyse the between-group differences in 3D dual-energy X-ray absorptiometry (DXA) parameters at the hip site among patients with acromegaly or adult growth hormone deficiency (AGHD) and a healthy control group. The current cross-sectional study includes data for 67 adults, 20 with acromegaly, 14 with AGHD and 33 healthy controls. We obtained the areal bone mineral density (aBMD) outcomes using DXA and cortical and trabecular parameters using 3D-DXA software (3D-SHAPER). The mean-adjusted 3D-DXA parameters did not differ between acromegaly patients and the controls (p > 0.05); however, we found cortical bone impairment (−7.3% to −8.4%; effect size (ES) = 0.78) in AGHD patients (p < 0.05). Differences in the cortical bone parameters were more evident when comparing AGHD patients (−8.5% to −16.2%; ES = 1.22 to 1.24) with acromegaly patients (p < 0.05). In brief, the 3D mapping highlighted the trochanter as the site with greater cortical bone differences between acromegaly patients and the controls. Overall, AGHD patients displayed lower cortical parameters at the trochanter, femoral neck and intertrochanter compared to the controls and acromegaly patients. To sum up, 3D-DXA provided useful information about the characteristics of bone involvement in growth hormone (GH)-related disorders. Patients with AGHD showed distinct involvement of the cortical structure.

Three-Dimensional Quantification of Bone Mineral Density in the Distal Femur and Proximal Tibia Based on Computed Tomography: In Vitro Evaluation of an Extended Standardization Method

While alterations in bone mineral density (BMD) are of interest in a number of musculoskeletal conditions affecting the knee, their analysis is limited by a lack of tools able to take full advantage of modern imaging modalities. This study introduced a new method, combining computed tomography (CT) and computational anatomy algorithms, to produce standardized three-dimensional BMD quantification in the distal femur and proximal tibia. The method was evaluated on ten cadaveric knees CT-scanned twice and processed following three different experimental settings to assess the influence of different scans and operators. The median reliability (intraclass correlation coefficient (ICC)) ranged from 0.96 to 0.99 and the median reproducibility (precision error (RMSSD)) ranged from 3.97 to 10.75 mg/cc for the different experimental settings. In conclusion, this paper presented a method to standardize three-dimensional knee BMD with excellent reliability and adequate reproducibility to be used in research and clinical applications. The perspectives offered by this novel method are further reinforced by the fact it relies on conventional CT scan of the knee. The standardization method introduced in this work is not limited to BMD and could be adapted to quantify other bone parameters in three dimension based on CT images or images acquired using different modalities.

Three-Dimensional Surface-Based Analysis of Cartilage MRI Data in Knee Osteoarthritis: Validation and Initial Clinical Application

BACKGROUND

Traditional quantitative analysis of cartilage with MRI averages measurements (eg, thickness) across regions-of-interest (ROIs) which may reduce responsiveness.

PURPOSE

To validate and describe clinical application of a semiautomated surface-based method for analyzing cartilage relaxation times ("composition") and morphology on MRI, 3D cartilage surface mapping (3D-CaSM).

STUDY TYPE

Validation study in cadaveric knees and prospective observational (cohort) study in human participants.

POPULATION

Four cadaveric knees and 14 participants aged 40-60 with mild-moderate knee osteoarthritis (OA) and 6 age-matched healthy volunteers, imaged at baseline, 1, and 6 months.

FIELD STRENGTH/SEQUENCE

3D spoiled gradient echo, T1 rho/T2 magnetization-prepared 3D fast spin echo for mapping of T1 rho/T2 relaxation times and delayed gadolinium enhanced MRI of cartilage (dGEMRIC) using variable flip angle T1 relaxation time mapping at 3T.

ASSESSMENT

3D-CaSM was validated against high-resolution peripheral quantitative computed tomography (HRpQCT) in cadaveric knees, with comparison to expert manual segmentation. The clinical study assessed test-retest repeatability and sensitivity to change over 6 months for cartilage thickness and relaxation times.

STATISTICAL TESTS

Bland-Altman analysis was performed for the validation study and evaluation of test-retest repeatability. Six-month changes were assessed via calculation of the percentage of each cartilage surface affected by areas of significant change (%SC), defined using thresholds based on area and smallest detectable difference (SDD).

RESULTS

Bias and precision (0.06 ± 0.25 mm) of 3D-CaSM against reference HRpQCT data were comparable to expert manual segmentation (-0.13 ± 0.26 mm). 3D-CaSM demonstrated significant (>SDD) 6-month changes in cartilage thickness and relaxation times in both OA participants and healthy controls. The parameter demonstrating the greatest 6-month change was T2 relaxation time (OA median %SC [IQR] = 8.8% [5.5 to 12.6]).

DATA CONCLUSION

This study demonstrates the construct validity and potential clinical utility of 3D-CaSM, which may offer advantages to conventional ROI-based methods.

LEVEL OF EVIDENCE

2.

TECHNICAL EFFICACY STAGE

2. J. Magn. Reson. Imaging 2020;52:1139-1151.

An experimental procedure to perform mechanical characterization of small-sized bone specimens from thin femoral cortical wall

The cortical shell of the femoral neck plays a role in determining the overall neck strength. However, there is a lack of knowledge about the mechanical properties of cortical tissue of the femoral neck due to challenges in implementing accurate testing protocols for the thin shell. Indeed, mechanical properties are commonly derived from mechanical testing performed on tissue samples extracted from the femoral diaphysis, i.e. assuming tissue homogeneity along the femur. The aim of this work was to set up a reliable methodology to determine mechanical properties of bone samples extracted from thin cortical shell of the femoral neck. A three-point bending test was used to determine elastic and post-elastic properties of cortical bone samples extracted from the inferior and superior femoral neck. An optical system was used to monitor the sample deflection. Accuracy was preliminarily evaluated by determining the elastic modulus of an aluminium alloy. A good intra- and inter-sample variability was found on determining aluminium elastic modulus: 1.6% and 3.6%, respectively. Additionally, aluminium elastic modulus value was underestimated by less than 1%. A pilot trial was performed on a human femoral neck to assess the procedure feasibility. A total of 22 samples were extracted from the inferior and superior femoral neck and successfully tested. Preliminary results suggest that mechanical properties of cortical bone tissue extracted from human femoral neck might be side dependent, the superior tissue seems to exhibit better mechanical properties than the inferior one, at least in terms of yield stress and maximum strain. This supposedly different mechanical competence must be further investigated. The proposed procedure makes it feasible to carry out such studies.

X-ray-based quantitative osteoporosis imaging at the spine

Osteoporosis is a metabolic bone disease with a high prevalence that affects the population worldwide, particularly the elderly. It is often due to fractures associated with bone fragility that the diagnosis of osteoporosis becomes clinically evident. However, early diagnosis would be necessary to initiate therapy and to prevent occurrence of further fractures, thus reducing morbidity and mortality. X-ray-based imaging plays a key role for fracture risk assessment and monitoring of osteoporosis. Whereas over decades dual-energy X-ray absorptiometry (DXA) has been the main method used and still reflects the reference standard, another modality reemerges with quantitative computed tomography (QCT) because of its three-dimensional advantages and the opportunistic exploitation of routine CT scans. Against this background, this article intends to review and evaluate recent advances in the field of X-ray-based quantitative imaging of osteoporosis at the spine. First, standard DXA with the recent addition of trabecular bone score (TBS) is presented. Secondly, standard QCT, dual-energy BMD quantification, and opportunistic BMD screening in non-dedicated CT exams are discussed. Lastly, finite element analysis and microstructural parameter analysis are reviewed.

Analysis of Bone Impairment by 3D DXA Hip Measures in Patients With Primary Hyperparathyroidism: A Pilot Study

CONTEXT

Primary hyperparathyroidism (PHPT) has been related to bone loss. Dual-energy x-ray absorptiometry (DXA) cannot distinguish between trabecular and cortical bone compartments but the recently developed three-dimensional (3D)-DXA software might overcome this issue.

OBJECTIVE

To examine the differences in DXA-derived areal bone mineral density (aBMD) and 3D-DXA parameters at the hip site between patients with PHPT and a healthy control group.

DESIGN

Cross-sectional pilot study.

SETTING

Hospital.

PATIENTS

80 adults (59.5 ± 9.1 yrs), 40 with PHPT and 40 age- and sex-matched healthy controls.

MEASURES

aBMD (g/cm2) of the femoral neck, trochanter, shaft, and total hip was assessed using DXA. Cortical surface (sBMD, mg/cm2), cortical volumetric BMD (vBMD, mg/cm3), trabecular vBMD (mg/cm3), integral vBMD (mg/cm3) and cortical thickness (mm) was assessed using 3D-DXA software.

RESULTS

Mean-adjusted values showed lower aBMD (7.5%-12.2%, effect size: 0.51-1.01) in the PHPT group compared with the control group (all P < 0.05). 3D-DXA revealed bone impairment (3.7%-8.5%, effect size: 0.47-0.65) in patients with PHPT, mainly in cortical parameters (all P < 0.05). However, differences in trabecular vBMD were not statistically significant (P = 0.055). The 3D mapping showed lower cortical sBMD, cortical vBMD, and cortical thickness at the trochanter and diaphysis in the PHPT group (P < 0.05) compared with the control group. In both groups, the presence of osteopenia or osteoporosis is related to lower cortical bone.

CONCLUSIONS

aBMD and cortical 3D parameters are impaired in patients with PHPT versus healthy controls. The vBMD of the trabecular compartment seems to be affected, although to a lesser extent.

Quantitative susceptibility mapping of the spine using in-phase echoes to initialize inhomogeneous field and R2* for the nonconvex optimization problem of fat-water separation

Quantitative susceptibility mapping (QSM) of human spinal vertebrae from a multi-echo gradient-echo (GRE) sequence is challenging, because comparable amounts of fat and water in the vertebrae make it difficult to solve the nonconvex optimization problem of fat-water separation (R2*-IDEAL) for estimating the magnetic field induced by tissue susceptibility. We present an in-phase (IP) echo initialization of R2*-IDEAL for QSM in the spinal vertebrae. Ten healthy human subjects were recruited for spine MRI. A 3D multi-echo GRE sequence was implemented to acquire out-phase and IP echoes. For the IP method, the R2* and field maps estimated by separately fitting the magnitude and phase of IP echoes were used to initialize gradient search R2*-IDEAL to obtain final R2*, field, water, and fat maps, and the final field map was used to generate QSM. The IP method was compared with the existing Zero method (initializing the field to zero), VARPRO-GC (variable projection using graphcuts but still initializing the field to zero), and SPURS (simultaneous phase unwrapping and removal of chemical shift using graphcuts for initialization) on both simulation and in vivo data. The single peak fat model was also compared with the multi-peak fat model. There was no substantial difference on QSM between the single peak and multi-peak fat models, but there were marked differences among different initialization methods. The simulations demonstrated that IP provided the lowest error in the field map. Compared to Zero, VARPRO-GC and SPURS, the proposed IP method provided substantially improved spine QSM in all 10 subjects.

Effects of inferior alveolar nerve rupture on bone remodeling of the mandible: A preliminary study

Although various animal studies have indicated that sensory nerves played an important role in bone metabolism and nerve injury could impair the process of bone remodeling, the actual effect of sensory nerve rupture on human bones remains unclear. The aim of this preliminary study was to investigate the effect of inferior alveolar nerve (IAN) rupture on mandibular bone remodeling of patients underwent bilateral sagittal split ramus osteotomy (BSSRO).Ten patients with unilateral IAN rupture during BSSRO were involved in this study. Neurosensory examinations were employed to assess the sensory function of bilateral IAN. The remodeling process of the post-operational mandible was evaluated by panoramic radiographs and computed tomography (CT) scans.Neurosensory examinations indicated that nerve rupture resulted in significant hypoesthesia at the IAN-rupture side. Assessment of panoramic radiographs showed no evident alterations of bone structure at the IAN-rupture side of mandible. Evaluation of CT images also indicated no statistical difference in bone density and thickness between IAN-rupture side and contralateral side.Accordingly, our study indicated that IAN rupture may not significantly impair the short-term bone remodeling process of human mandible.

Comparing rib cortical thickness measurements from computed tomography (CT) and Micro-CT

BACKGROUND

The objective of this study was to compare cortical thickness of rib specimens scanned with clinical computed tomography (clinical-CT) at 0.5 and 1.0 mm slice thickness versus micro-CT at 0.05 mm slice thickness. Cortical thickness variation and accuracy was explored by anatomical region (anterior vs. lateral) and cross-sectional quadrants (superior, interior, inferior, and exterior).

METHODS

A validated cortical thickness algorithm was applied to clinical-CT and micro-CT scans of 17 rib specimens from six male post mortem human subjects aged 42-81 years. Each rib specimen was segmented and the thickness measurements were partitioned into cross-sectional quadrants in the anterior and lateral regions of the rib. Within each rib quadrant, the following were calculated: average thickness ± standard deviation, mean thickness difference between clinical-CT and micro-CT, and a thickness ratio between clinical-CT and micro-CT. Correlations from linear regression and paired-t tests were determined for paired clinical-CT and micro-CT results.

RESULTS

On average, the 0.5 mm clinical-CT underestimated the micro-CT thickness by 0.005 mm, while the 1.0 mm clinical-CT overestimated the micro-CT thickness by 0.149 mm. Thickness derived from 0.5 mm clinical-CT showed greater significant linear correlations (p < 0.05) with micro-CT thickness compared to 1.0 mm clinical-CT.

CONCLUSIONS

The small mean differences and thickness ratios near 1 show validation for the cortical thickness algorithm when applied to rib clinical-CT scans. Using clinical-CT scans as way to accurately measure rib cortical thickness offers a non-invasive way to analyze millions of CT scans collected each year from males and females of all ages.