A Review of CT-Based Fracture Risk Assessment with Finite Element Modeling and Machine Learning

DXA-based 3D finite element models predict hip fractures better than areal BMD in elderly women

Bone strength is a major contributor to fracture risk. Areal bone mineral density (aBMD) obtained from dual-energy X-ray absorptiometry (DXA) is used as a surrogate for bone strength in fracture risk prediction. 3D finite element (FE) models predict bone strength better than aBMD but need 3D computed tomography and are not automated. We have earlier developed a method to automatically reconstruct the 3D hip anatomy from a 2D hip DXA image, followed by subject-specific FE-based prediction of proximal femoral strength. In this study, we evaluate the method's ability to predict incident hip fractures in a population-based cohort of women (OSTPRE). We used a sub-cohort including 46 cases with a hip fracture (<10 years from DXA scan) and 2 healthy controls to each hip fracture case, matched by age, height, and body mass index. We automatically reconstructed the 3D hip anatomy and predicted proximal femoral strength using FE analysis for all the subjects of the sub-cohort. The FE-predicted proximal femoral strength was a significantly better predictor of incident hip fractures than aBMD (difference in area under the receiver operating characteristics curve, ΔAUROC = 0.10). This is the first time that 3D FE models obtained from a 2D hip DXA scan outperform aBMD in predicting incident hip fractures in a population-based prospectively followed cohort of women. Our approach provided an improved fracture risk prediction in a clinically feasible manner (only one single DXA image is needed) and without additional costs compared to the current clinical approach.

Faster R-CNN model for target recognition and diagnosis of scapular fractures

Objective

This study aims to establish a diagnostic model for scapular fractures using a convolutional neural network (CNN) and to discuss the clinical advantages of this model in diagnosing such complex conditions.

Methods

Computed tomography (CT) images of 90 patients with scapular fractures were collected. A faster R-CNN-based recognition model was developed and compared with manual diagnosis. External validation was conducted to evaluate the model's accuracy, sensitivity, specificity, positive predictive value, and negative predictive value.

Results

The CNN model, when combined with medical expert interpretation, demonstrated significantly higher specificity and positive predictive value compared to orthopedist-independent interpretation and algorithm-independent prediction (P < 0.05). The area under the curve (AUC) value of the combined approach was significantly higher than that of orthopedist-independent interpretation and algorithm-independent prediction groups, with statistically significant differences (P < 0.05). The accuracy of the CNN algorithm model combined with orthopedist interpretation was 97.78 %, significantly higher than orthopedist-independent interpretation (82.95 %) and CNN algorithm-independent prediction (92.05 %) (P < 0.05).

Conclusions

The CNN-based recognition model for scapular fractures can assist clinicians in improving their diagnostic accuracy and precision in identifying such fractures on CT images.

Gender differences in L1 vertebral strength in adults 50+ using automated CT-based finite element analysis

Osteoporosis is usually diagnosed using a Bone Mineral Density test using dual-energy X-ray Absorptiometry. However, it is limited by low testing rates and the inability to directly measure bone strength. Finite Element Analysis allows for a more detailed assessment of bone strength. However, its modeling complexity and high computational time requirements pose challenges. This study aims to develop customized MATLAB programs to automate the creation of heterogeneous bone models, streamlining preprocessing to reduce time, computational costs, and minimize variability from manual processes. The focus is on establishing a prediction model for the structural strength of the L1 vertebral body using patient-specific CT data, thereby aiding in the prediction of vertebral fracture risk. The CT images are stacked into a 3D array, and the pixel values are converted by Hounsfield units based on CT image. The bone segment and elasticity values are established based on the Hounsfield units. After modeling, strain and stress analysis were performed through the solver LS-DYNA. The compression force was distributed vertically on the upper endplate of the vertebral body. All nodes in the subvertebral plane were fully constrained. For comparison, vertebral models were automatically established and analyzed from recruited subjects. This study collected spine CT imaging datasets from 52 subjects, comprising 28 males and 24 females aged between 50 and 95 years. Preprocessing and mechanical analysis for each subject took an average of approximately 579.6 seconds. Analysis of the results indicated that women over 50 years of age exhibited higher strain and stress values in their vertebral models compared to men under the same applied force, highlighting gender-specific differences in biomechanical characteristics. This study effectively employed a practical approach to identify and select specific spinal segments from CT images, facilitating the automated creation of 3D models for subsequent finite element analysis. The predictive model generated results consistent with previous studies involving mechanical testing on actual human bones. Notably, the implementation of our predictive model substantially decreased processing time for Finite Element Analysis, rendering it more suitable for clinical use and easier to extend for future application.

Ensuring anatomical integrity and shared contact surfaces in vertebra and disc models: a segmentation-based smoothing approach

Due to limited MRI resolution, patient-specific simulation models derived from medical images often lack bio-fidelity. To address this, we present a smoothing pipeline for generating high-fidelity meshes of vertebrae and intervertebral discs from medical images, which serve as a base for biomechanical simulations. Using a diverse array of vertebrae smoothing algorithms, including e.g. Laplace and Taubin, we systematically explored 136 combinations across six protocols. Subsequently, an adaptive smoothing algorithm was developed for intervertebral disc meshes. By adjusting vertex locations to those of the vertebra mesh, we ensured seamless alignment of contact surfaces, including shared nodes. Evaluation of our pipeline against conventional smoothing methods demonstrates superior edge preservation and reduced stair-step effects, enhancing the fidelity of the generated meshes. Finite Element Method simulations further confirmed the accuracy of our selective smoothing pipeline, showing increased notch stress. Validated on a diverse dataset, our smoothing pipeline generates patient-specific models with enhanced biomechanical fidelity, enabling large-scale studies and biomechanical insights into spine pathologies.

From MRI to FEM: an automated pipeline for biomechanical simulations of vertebrae and intervertebral discs

Introduction

Biomechanical simulations can enhance our understanding of spinal disorders. Applied to large cohorts, they can reveal complex mechanisms beyond conventional imaging. Therefore, automating the patient-specific modeling process is essential.

Methods

We developed an automated and robust pipeline that generates and simulates biofidelic vertebrae and intervertebral disc finite element method (FEM) models based on automated magnetic resonance imaging (MRI) segmentations. In a first step, anatomically-constrained smoothing approaches were implemented to ensure seamless contact surfaces between vertebrae and discs with shared nodes. Subsequently, surface meshes were filled isotropically with tetrahedral elements. Lastly, simulations were executed. The performance of our pipeline was evaluated using a set of 30 patients from an in-house dataset that comprised an overall of 637 vertebrae and 600 intervertebral discs. We rated mesh quality metrics and processing times.

Results

With an average number of 21 vertebrae and 20 IVDs per subject, the average processing time was 4.4 min for a vertebra and 31 s for an IVD. The average percentage of poor quality elements stayed below 2% in all generated FEM models, measured by their aspect ratio. Ten vertebra and seven IVD FE simulations failed to converge.

Discussion

The main goal of our work was to automate the modeling and FEM simulation of both patient-specific vertebrae and intervertebral discs with shared-node surfaces directly from MRI segmentations. The biofidelity, robustness and time-efficacy of our pipeline marks an important step towards investigating large patient cohorts for statistically relevant, biomechanical insight.

Explainable machine-learning-based prediction of QCT/FEA-calculated femoral strength under stance loading configuration using radiomics features

Finite element analysis can provide precise femoral strength assessment. However, its modeling procedures were complex and time-consuming. This study aimed to develop a model to evaluate femoral strength calculated by quantitative computed tomography-based finite element analysis (QCT/FEA) under stance loading configuration, offering an effective, simple, and explainable method. One hundred participants with hip QCT images were selected from the Hong Kong part of the Osteoporotic fractures in men cohort. Radiomics features were extracted from QCT images. Filter method, Pearson correlation analysis, and least absolute shrinkage and selection operator method were employed for feature selection and dimension reduction. The remaining features were utilized as inputs, and femoral strengths were calculated as the ground truth through QCT/FEA. Support vector regression was applied to develop a femoral strength prediction model. The influence of various numbers of input features on prediction performance was compared, and the femoral strength prediction model was established. Finally, Shapley additive explanation, accumulated local effects, and partial dependency plot methods were used to explain the model. The results indicated that the model performed best when six radiomics features were selected. The coefficient of determination (R2), the root mean square error, the normalized root mean square error, and the mean squared error on the testing set were 0.820, 1016.299 N, 10.645%, and 750.827 N, respectively. Additionally, these features all positively contributed to femoral strength prediction. In conclusion, this study provided a noninvasive, effective, and explainable method of femoral strength assessment, and it may have clinical application potential.

CT image-based biomarkers for opportunistic screening of osteoporotic fractures: a systematic review and meta-analysis

The use of opportunistic computed tomography (CT) image-based biomarkers may be a low-cost strategy for screening older individuals at high risk for osteoporotic fractures and populations that are not sufficiently targeted. This review aimed to assess the discriminative ability of image-based biomarkers derived from existing clinical routine CT scans for hip, vertebral, and major osteoporotic fracture prediction. A systematic search in PubMed MEDLINE, Embase, Cochrane, and Web of Science was conducted from the earliest indexing date until July 2023. The evaluation of study quality was carried out using a modified Quality Assessment Tool for Diagnostic Accuracy Studies (QUADAS-2) checklist. The primary outcome of interest was the area under the curve (AUC) and its corresponding 95% confidence intervals (CIs) obtained for four main categories of biomarkers: areal bone mineral density (BMD), image attenuation, volumetric BMD, and finite element (FE)-derived biomarkers. The meta-analyses were performed using random effects models. Sixty-one studies were included in this review, among which 35 were synthesized in a meta-analysis and the remaining articles were qualitatively synthesized. In comparison to the pooled AUC of areal BMD (0.73 [95% CI 0.71-0.75]), the pooled AUC values for predicting osteoporotic fractures for FE-derived parameters (0.77 [95% CI 0.72-0.81]; p < 0.01) and volumetric BMD (0.76 [95% CI 0.71-0.81]; p < 0.01) were significantly higher, but there was no significant difference with the pooled AUC for image attenuation (0.73 [95% CI 0.66-0.79]; p = 0.93). Compared to areal BMD, volumetric BMD and FE-derived parameters may provide a significant improvement in the discrimination of osteoporotic fractures using opportunistic CT assessments.

DXA-based statistical models of shape and intensity outperform aBMD hip fracture prediction: A retrospective study

Areal bone mineral density (aBMD) currently represents the clinical gold standard for hip fracture risk assessment. Nevertheless, it is characterised by a limited prediction accuracy, as about half of the people experiencing a fracture are not classified as at being at risk by aBMD. In the context of a progressively ageing population, the identification of accurate predictive tools would be pivotal to implement preventive actions. In this study, DXA-based statistical models of the proximal femur shape, intensity (i.e., density) and their combination were developed and employed to predict hip fracture on a retrospective cohort of post-menopausal women. Proximal femur shape and pixel-by-pixel aBMD values were extracted from DXA images and partial least square (PLS) algorithm adopted to extract corresponding modes and components. Subsequently, logistic regression models were built employing the first three shape, intensity and shape-intensity PLS components, and their ability to predict hip fracture tested according to a 10-fold cross-validation procedure. The area under the ROC curves (AUC) for the shape, intensity, and shape-intensity-based predictive models were 0.59 (95%CI 0.47-0.69), 0.80 (95%CI 0.70-0.90) and 0.83 (95%CI 0.73-0.90), with the first being significantly lower than the latter two. aBMD yielded an AUC of 0.72 (95%CI 0.59-0.82), found to be significantly lower than the shape-intensity-based predictive model. In conclusion, a methodology to assess hip fracture risk uniquely based on the clinically available imaging technique, DXA, is proposed. Our study results show that hip fracture risk prediction could be enhanced by taking advantage of the full set of information DXA contains.

Development and validation of a semi-automated and unsupervised method for femur segmentation from CT

Quantitative computed tomography (QCT)-based in silico models have demonstrated improved accuracy in predicting hip fractures with respect to the current gold standard, the areal bone mineral density. These models require that the femur bone is segmented as a first step. This task can be challenging, and in fact, it is often almost fully manual, which is time-consuming, operator-dependent, and hard to reproduce. This work proposes a semi-automated procedure for femur bone segmentation from CT images. The proposed procedure is based on the bone and joint enhancement filter and graph-cut algorithms. The semi-automated procedure performances were assessed on 10 subjects through comparison with the standard manual segmentation. Metrics based on the femur geometries and the risk of fracture assessed in silico resulting from the two segmentation procedures were considered. The average Hausdorff distance (0.03 ± 0.01 mm) and the difference union ratio (0.06 ± 0.02) metrics computed between the manual and semi-automated segmentations were significantly higher than those computed within the manual segmentations (0.01 ± 0.01 mm and 0.03 ± 0.02). Besides, a blind qualitative evaluation revealed that the semi-automated procedure was significantly superior (p < 0.001) to the manual one in terms of fidelity to the CT. As for the hip fracture risk assessed in silico starting from both segmentations, no significant difference emerged between the two (R2 = 0.99). The proposed semi-automated segmentation procedure overcomes the manual one, shortening the segmentation time and providing a better segmentation. The method could be employed within CT-based in silico methodologies and to segment large volumes of images to train and test fully automated and supervised segmentation methods.

Fully automated CT imaging biomarkers for opportunistic prediction of future hip fractures

OBJECTIVE

Assess automated CT imaging biomarkers in patients who went on to hip fracture, compared with controls.

METHODS

In this retrospective case-control study, 6926 total patients underwent initial abdominal CT over a 20-year interval at one institution. A total of 1308 patients (mean age at initial CT, 70.5 ± 12.0 years; 64.4% female) went on to hip fracture (mean time to fracture, 5.2 years); 5618 were controls (mean age 70.3 ± 12.0 years; 61.2% female; mean follow-up interval 7.6 years). Validated fully automated quantitative CT algorithms for trabecular bone attenuation (at L1), skeletal muscle attenuation (at L3), and subcutaneous adipose tissue area (SAT) (at L3) were applied to all scans. Hazard ratios (HRs) comparing highest to lowest risk quartiles and receiver operating characteristic (ROC) curve analysis including area under the curve (AUC) were derived.

RESULTS

Hip fracture HRs (95% CI) were 3.18 (2.69-3.76) for low trabecular bone HU, 1.50 (1.28-1.75) for low muscle HU, and 2.18 (1.86-2.56) for low SAT. 10-year ROC AUC values for predicting hip fracture were 0.702, 0.603, and 0.603 for these CT-based biomarkers, respectively. Multivariate combinations of these biomarkers further improved predictive value; the 10-year ROC AUC combining bone/muscle/SAT was 0.733, while combining muscle/SAT was 0.686.

CONCLUSION

Opportunistic use of automated CT bone, muscle, and fat measures can identify patients at higher risk for future hip fracture, regardless of the indication for CT imaging.

ADVANCES IN KNOWLEDGE

CT data can be leveraged opportunistically for further patient evaluation, with early intervention as needed. These novel AI tools analyse CT data to determine a patient's future hip fracture risk.

The effect of deep learning-based lesion segmentation on failure load calculations of metastatic femurs using finite element analysis

Bone ranks as the third most frequent tissue affected by cancer metastases, following the lung and liver. Bone metastases are often painful and may result in pathological fracture, which is a major cause of morbidity and mortality in cancer patients. To quantify fracture risk, finite element (FE) analysis has shown to be a promising tool, but metastatic lesions are typically not specifically segmented and therefore their mechanical properties may not be represented adequately. Deep learning methods potentially provide the opportunity to automatically segment these lesions and change the mechanical properties more adequately. In this study, our primary focus was to gain insight into the performance of an automatic segmentation algorithm for femoral metastatic lesions using deep learning methods and the subsequent effects on FE outcomes. The aims were to determine the similarity between manual segmentation and automatic segmentation; the differences in predicted failure load between FE models with automatically segmented osteolytic and mixed lesions and the models with CT-based lesion values (the gold standard); and the effect on the BOne Strength (BOS) score (failure load adjusted for body weight) and subsequent fracture risk assessments. From two patient cohorts, a total number of 50 femurs with osteolytic and mixed metastatic lesions were included in this study. The femurs were segmented from CT images and transferred into FE meshes. The material behavior was implemented as non-linear isotropic. These FE models were considered as gold standard (Finite Element no Segmented Lesion: FE-no-SL), whereby the local calcium equivalent density of both femur and metastatic lesion was extracted from CT-values. Lesions in the femur were manually segmented by two biomechanical experts after which final lesion segmentation for each femur was obtained based on consensus of opinions between two observers. Subsequently, a self-configuring variant of the popular deep learning model U-Net known as nnU-Net was used to automatically segment metastatic lesions within the femur. For these models with segmented lesions (Finite Element with Segmented Lesion: FE-with-SL), the calcium equivalent density within the metastatic lesions was set to zero after being segmented by the neural network, simulating absence of load-bearing capacity of these lesions. The models (either with or without automatically segmented lesions) were loaded incrementally in axial direction until failure was simulated. Dice coefficient was used to evaluate the similarity of the manual and automatic segmentation. Mean calcium equivalent density values within the automatically segmented lesions were calculated. Failure loads and patterns were determined. Furthermore, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for both groups by comparing the predictions to the occurrence or absence of actual fracture within the patient cohorts. The automatic segmentation algorithm performed in a none-robust manner. Dice coefficients describing the similarity between consented manual and automatic segmentations were relatively low (mean 0.45 ± standard deviation 0.33, median 0.54). Failure load difference between the FE-no-SL and FE-with-SL groups varied from 0 % to 48 % (mean 6.6 %). Correlation analysis of failure loads between the two groups showed a strong relationship (R2 > 0.9). From the 50 cases, four cases showed clear deviations for which models with automatic lesion segmentation (FE-with-SL) showed considerably lower failure loads. In the whole database including osteolytic and mixed lesions, sensitivity and NPV remained the same, but specificity and PPV decreased from 94 % to 83 %, and from 78 % to 54 % respectively from FE-no-SL to FE-with-SL. This study indicates that the nnU-Net yielded none-robust outcomes in femoral lesion segmentation and that other segmentation algorithms should be considered. However, the difference in failure pattern and failure load between FE models with automatically segmented osteolytic and mixed lesions were relatively small in most cases with a few exceptions. On the other hand, the accuracy of fracture risk assessment using the BOS score was lower compared to the FE-no-SL. In conclusion, this study showed that automatic lesion segmentation is a none-solved issue and therefore, quantifying lesion characteristics and the subsequent effect on the fracture risk using deep learning will remain challenging.

Quantitative Skeletal Imaging and Image-Based Modeling in Pediatric Orthopaedics

PURPOSE OF REVIEW

Musculoskeletal imaging serves a critical role in clinical care and orthopaedic research. Image-based modeling is also gaining traction as a useful tool in understanding skeletal morphology and mechanics. However, there are fewer studies on advanced imaging and modeling in pediatric populations. The purpose of this review is to provide an overview of recent literature on skeletal imaging modalities and modeling techniques with a special emphasis on current and future uses in pediatric research and clinical care.

RECENT FINDINGS

While many principles of imaging and 3D modeling are relevant across the lifespan, there are special considerations for pediatric musculoskeletal imaging and fewer studies of 3D skeletal modeling in pediatric populations. Improved understanding of bone morphology and growth during childhood in healthy and pathologic patients may provide new insight into the pathophysiology of pediatric-onset skeletal diseases and the biomechanics of bone development. Clinical translation of 3D modeling tools developed in orthopaedic research is limited by the requirement for manual image segmentation and the resources needed for segmentation, modeling, and analysis. This paper highlights the current and future uses of common musculoskeletal imaging modalities and 3D modeling techniques in pediatric orthopaedic clinical care and research.

Characterizing Bone Phenotypes Related to Skeletal Fragility Using Advanced Medical Imaging

PURPOSE OF REVIEW

Summarize the recent literature that investigates how advanced medical imaging has contributed to our understanding of skeletal phenotypes and fracture risk across the lifespan.

RECENT FINDINGS

Characterization of bone phenotypes on the macro-scale using advanced imaging has shown that while wide bones are generally stronger than narrow bones, they may be more susceptible to age-related declines in bone strength. On the micro-scale, HR-pQCT has been used to identify bone microarchitecture phenotypes that improve stratification of fracture risk based on phenotype-specific risk factors. Adolescence is a key phase for bone development, with distinct sex-specific growth patterns and significant within-sex bone property variability. However, longitudinal studies are needed to evaluate how early skeletal growth impacts adult bone phenotypes and fracture risk. Metabolic and rare bone diseases amplify fracture risk, but the interplay between bone phenotypes and disease remains unclear. Although bone phenotyping is a promising approach to improve fracture risk assessment, the clinical availability of advanced imaging is still limited. Consequently, alternative strategies for assessing and managing fracture risk include vertebral fracture assessment from clinically available medical imaging modalities/techniques or from fracture risk assessment tools based on clinical risk factors. Bone fragility is not solely determined by its density but by a combination of bone geometry, distribution of bone mass, microarchitecture, and the intrinsic material properties of bone tissue. As such, different individuals can exhibit distinct bone phenotypes, which may predispose them to be more vulnerable or resilient to certain perturbations that influence bone strength.

3D Finite Element Models Reconstructed From 2D Dual-Energy X-Ray Absorptiometry (DXA) Images Improve Hip Fracture Prediction Compared to Areal BMD in Osteoporotic Fractures in Men (MrOS) Sweden Cohort

Bone strength is an important contributor to fracture risk. Areal bone mineral density (aBMD) derived from dual-energy X-ray absorptiometry (DXA) is used as a surrogate for bone strength in fracture risk prediction tools. 3D finite element (FE) models predict bone strength better than aBMD, but their clinical use is limited by the need for 3D computed tomography and lack of automation. We have earlier developed a method to reconstruct the 3D hip anatomy from a 2D DXA image, followed by subject-specific FE-based prediction of proximal femoral strength. In the current study, we aim to evaluate the method's ability to predict incident hip fractures in a population-based cohort (Osteoporotic Fractures in Men [MrOS] Sweden). We defined two subcohorts: (i) hip fracture cases and controls cohort: 120 men with a hip fracture (<10 years from baseline) and two controls to each hip fracture case, matched by age, height, and body mass index; and (ii) fallers cohort: 86 men who had fallen the year before their hip DXA scan was acquired, 15 of which sustained a hip fracture during the following 10 years. For each participant, we reconstructed the 3D hip anatomy and predicted proximal femoral strength in 10 sideways fall configurations using FE analysis. The FE-predicted proximal femoral strength was a better predictor of incident hip fractures than aBMD for both hip fracture cases and controls (difference in area under the receiver operating characteristics curve, ΔAUROC = 0.06) and fallers (ΔAUROC = 0.22) cohorts. This is the first time that FE models outperformed aBMD in predicting incident hip fractures in a population-based prospectively followed cohort based on 3D FE models obtained from a 2D DXA scan. Our approach has potential to notably improve the accuracy of fracture risk predictions in a clinically feasible manner (only one single DXA image is needed) and without additional costs compared to the current clinical approach. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).