Opportunistic diagnosis of osteoporosis, fragile bone strength and vertebral fractures from routine CT scans; a review of approved technology systems and pathways to implementation

Evaluation and analysis of risk factors for adverse events of the fractured vertebra post-percutaneous kyphoplasty: a retrospective cohort study using multiple machine learning models

BACKGROUND

Adverse events of the fractured vertebra (AEFV) post-percutaneous kyphoplasty (PKP) can lead to recurrent pain and neurological damage, which considerably affect the prognosis of patients and the quality of life. This study aimed to analyze the risk factors of AEFV and develop and select the optimal risk prediction model for AEFV to provide guidance for the prevention of this condition and enhancement of clinical outcomes.

METHODS

This work included 383 patients with primary osteoporotic vertebral compression fracture (OVCF) who underwent PKP. The patients were grouped based on the occurrence of AEFV postsurgery, and data were collected. Group comparisons and correlation analysis were conducted to identify potential risk factors, which were then included in the five prediction models. The performance indicators served as basis for the selection of the best model.

RESULTS

Multivariate logistic regression analysis revealed the following independent risk factors for AEFV: kissing spine (odds ratio (OR) = 8.47, 95% confidence interval (CI) 1.46-49.02), high paravertebral muscle fat infiltration grade (OR = 29.19, 95% CI 4.83-176.04), vertebral body computed tomography value (OR = 0.02, 95% CI 0.003-0.13, P < 0.001), and large Cobb change (OR = 5.31, 95% CI 1.77-15.77). The support vector machine (SVM) model exhibited the best performance in the prediction of the risk of AEFV.

CONCLUSION

Four independent risk factors were identified of AEFV, and five risk prediction models that can aid clinicians in the accurate identification of high-risk patients and prediction of the occurrence of AEFV were developed.

Opportunistic CT screening demonstrates increased risk for peri-articular fractures in osteoporotic patients

BACKGROUND

Underdiagnosis or undertreatment of osteoporosis consequently impacts individual morbidity and mortality, as well as on healthcare systems and communities as a whole. Dual-energy x-ray absorptiometry (DXA) is the gold standard method for identifying osteoporosis, however, opportunistic CT screening is capable of precisely estimating bone mineral density (BMD) in abdominopelvic imaging with no additional cost, radiation exposure or inconvenience to patients. This study uses opportunistic CT screening to determine the prevalence of osteoporosis and anatomic distribution patterns in patients presenting with lower extremity fractures at our institution.

HYPOTHESIS

Trauma patients with low bone mineral density (BMD) are more likely to present with peri-articular versus shaft fractures.

PATIENTS AND METHODS

We conducted a retrospective review of 721 patients presenting as trauma activations to the emergency department (ED) of a Level 1 Trauma Center with lower extremity fractures. Patients were excluded if under the age of 18 or lacking a CT scan upon arrival in the ED. Hounsfield Units (HU) were measured at the L1 vertebral level on CT scans to determine bone mineral density. Values of ≤100 HU were consistent with osteoporosis, whereas 101-150 HU were consistent with osteopenia.

RESULTS

The final cohort included 416 patients, with mean age of 49 ± 21 years. Average bone density was 203.9 ± 73.4 HU. 15.9% of patients were diagnosed as osteopenic and 9.9% as osteoporotic. 64.2% of fractures were peri-articular, 25.7% were shaft, and 10.1% were a combination. Peri-articular fractures were significantly more likely to have lower average BMD than shaft fractures (189 ± 74.7 HU vs. 230.6 ± 66.1 HU, p < 0.001).

DISCUSSION

Our study demonstrates a significant relationship between low bone mineral density and lower extremity fracture pattern, however, likely influenced by other factors such as sex. Opportunistic CT screening for osteoporosis in trauma settings provides ample opportunity for early detection of low BMD and implementation of highly effective lifestyle modification and pharmacotherapy intervention. Reduction in the overall incidence of peri-articular fracture with widespread adoption of opportunistic CT screening may lessen the morbidity, mortality, and total cost currently afflicting patients, healthcare systems, and communities.

LEVEL OF EVIDENCE

III, therapeutic.

The potential role for artificial intelligence in fracture risk prediction

Osteoporotic fractures are a major health challenge in older adults. Despite the availability of safe and effective therapies for osteoporosis, these therapies are underused in individuals at high risk for fracture, calling for better case-finding and fracture risk assessment strategies. Artificial intelligence (AI) and machine learning (ML) hold promise for enhancing identification of individuals at high risk for fracture by distilling useful features from high-dimensional data derived from medical records, imaging, and wearable devices. AI-ML could enable automated opportunistic screening for vertebral fractures and osteoporosis, home-based monitoring and intervention targeting lifestyle factors, and integration of multimodal features to leverage fracture prediction, ultimately aiding improved fracture risk assessment and individualised treatment. Optimism must be balanced with consideration for the explainability of AI-ML models, biases (including information inequity in numerically under-represented populations), model limitations, and net clinical benefit and workload impact. Clinical integration of AI-ML algorithms has the potential to transform osteoporosis management, offering a more personalised approach to reduce the burden of osteoporotic fractures.

Diagnostic Value of Hounsfield Units for Osteoporotic Thoracolumbar Vertebral Non-Compression Fractures in Elderly Patients with Low-Energy Injuries

Background

Thoracolumbar vertebral fractures are common pathological fractures caused by osteoporosis in the elderly. These fractures are challenging to detect. This study aimed to evaluate the diagnostic value of Hounsfield units for osteoporotic thoracolumbar vertebral non-compression fractures in elderly patients with low-energy fractures.

Methods

The retrospective case-control study included elderly patients diagnosed with osteoporotic thoracolumbar vertebral fractures and non-fractured patients who underwent computed tomography examinations for lumbar vertebra issues during July 2017 and June 2020.

Results

This study included 216 patients with fractures (38 males and 178 females; average age: 77.28±8.68 years) and 124 patients without fractures (21 males and 103 females; average age: 75.35±9.57 years). The difference in Hounsfield units of the target (intermediate) vertebral body significantly differed between the two groups (54.74 ± 21.84 vs 5.86 ± 5.14; p<0.001). The ratios of Hounsfield units were also significantly different between the two groups (1.38 ± 1.60 vs 0.13 ± 0.23; p<0.001). The cut-off value for the difference in Hounsfield units to detect osteoporotic spine fractures was 25.35, with high sensitivity (98.5%), specificity (99.9%), and the area under the curve (AUC) (0.999, 95% CI: 0.999-1). The cut-off value for the odds ratio of Hounsfield units was 0.260, with high sensitivity (99.1%), specificity (92.7%), and AUC (0.970, 95% CI: 0.949-0.992).

Conclusion

The difference between Hounsfield units and the odds ratio of Hounsfield units might help diagnose osteoporotic thoracolumbar vertebral non-compression fractures in elderly patients with low-energy fractures.

Trabecular bone mineral density as measured by thoracic vertebrae predicts incident hip and vertebral fractures: the multi-ethnic study of atherosclerosis

We evaluated the relationship of bone mineral density (BMD) by computed tomography (CT), to predict fractures in a multi-ethnic population. We demonstrated that vertebral and hip fractures were more likely in those patients with low BMD. This is one of the first studies to demonstrate that CT BMD derived from thoracic vertebrae can predict future hip and vertebral fractures.

PURPOSE/INTRODUCTION

Osteoporosis affects an enormous number of patients, of all races and both sexes, and its prevalence increases as the population ages. Few studies have evaluated the association between the vertebral trabecular bone mineral density(vBMD) and osteoporosis-related hip fracture in a multiethnic population, and no studies have demonstrated the predictive value of vBMD for fractures.

METHOD

We sought to determine the predictive value of QCT-based trabecular vBMD of thoracic vertebrae derived from coronary artery calcium scan for hip fractures in the Multi-Ethnic Study of Atherosclerosis(MESA), a nationwide multicenter cohort included 6814 people from six medical centers across the USA and assess if low bone density by QCT can predict future fractures. Measures were done using trabecular bone measures, adjusted for individual patients, from three consecutive thoracic vertebrae (BDI Inc, Manhattan Beach CA, USA) from non-contrast cardiac CT scans.

RESULTS

Six thousand eight hundred fourteen MESA baseline participants were included with a mean age of 62.2 ± 10.2 years, and 52.8% were women. The mean thoracic BMD is 162.6 ± 46.8 mg/cm3 (95% CI 161.5, 163.7), and 27.6% of participants (n = 1883) had osteoporosis (T-score 2.5 or lower). Over a median follow-up of 17.4 years, Caucasians have a higher rate of vertebral fractures (6.9%), followed by Blacks (4.4%), Hispanics (3.7%), and Chinese (3.0%). Hip fracture patients had a lower baseline vBMD as measured by QCT than the non-hip fracture group by 13.6 mg/cm3 [P < 0.001]. The same pattern was seen in the vertebral fracture population, where the mean BMD was substantially lower 18.3 mg/cm3 [P < 0.001] than in the non-vertebral fracture population. Notably, the above substantial relationship was unaffected by age, gender, race, BMI, hypertension, current smoking, medication use, or activity. Patients with low trabecular BMD of thoracic vertebrae showed a 1.57-fold greater risk of first hip fracture (HR 1.57, 95% CI 1.38-1.95) and a nearly threefold increased risk of first vertebral fracture (HR 2.93, 95% CI 1.87-4.59) compared to normal BMD patients.

CONCLUSION

There is significant correlation between thoracic trabecular BMD and the incidence of future hip and vertebral fracture. This study demonstrates that thoracic vertebrae BMD, as measured on cardiac CT (QCT), can predict both hip and vertebral fractures without additional radiation, scanning, or patient burden. Osteopenia and osteoporosis are markedly underdiagnosed. Finding occult disease affords the opportunity to treat the millions of people undergoing CT scans every year for other indications.

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.

Application of radiomics model based on lumbar computed tomography in diagnosis of elderly osteoporosis

A metabolic bone disease characterized by decreased bone formation and increased bone resorption is osteoporosis. It can cause pain and fracture of patients. The elderly are prone to osteoporosis and are more vulnerable to osteoporosis. In this study, radiomics are extracted from computed tomography (CT) images to screen osteoporosis in the elderly. Collect the plain scan CT images of lumbar spine, cut the region of interest of the image and extract radiomics features, use Lasso regression to screen variables and adjust complexity, use python language to model random forests, support vector machines, K nearest neighbor, and finally use receiver operating characteristic curve to evaluate the performance of the model, including precision, recall, accuracy and area under the curve (AUC). For the model, 14 radiolomics features were selected. The diagnosis performance of random forest model and support vector machine is good, all around 0.9. The AUC of K nearest neighbor model in training set and test set is 0.828 and 0.796, respectively. We selected the plain scan CT images of the elderly lumbar spine to build radiomics features model, which has good diagnostic performance and can be used as a tool to assist the diagnosis of osteoporosis in the elderly.

Femoral osteoporosis prediction model using autosegmentation and machine learning analysis with PyRadiomics on abdomen-pelvic computed tomography (CT)

Background

With the advancement of artificial intelligence technology and radiomics analysis, opportunistic prediction of osteoporosis with computed tomography (CT) is a new paradigm in osteoporosis screening. This study aimed to assess the diagnostic performance of osteoporosis prediction by the combination of autosegmentation of the proximal femur and machine learning analysis with a reference standard of dual-energy X-ray absorptiometry (DXA).

Methods

Abdomen-pelvic CT scans were retrospectively analyzed from 1,122 patients who received both DXA and abdomen-pelvic computed tomography (APCT) scan from January 2018 to December 2020. The study cohort consisted of a training cohort and a temporal validation cohort. The left proximal femur was automatically segmented, and a prediction model was built by machine-learning analysis using a random forest (RF) analysis and 854 PyRadiomics features. The technical success rate of autosegmentation, diagnostic test, area under the receiver operator characteristics curve (AUC), and precision recall curve (AUC-PR) analysis were used to analyze the training and validation cohorts.

Results

The osteoporosis prevalence of the training and validation cohorts was 24.5%, and 10.3%, respectively. The technical success rate of autosegmentation of the proximal femur was 99.7%. In the diagnostic test, the training and validation cohorts showed 78.4% vs. 63.3% sensitivity, 89.4% vs. 98.1% specificity. The prediction performance to identify osteoporosis within the groups used for training and validation cohort was high and the AUC and AUC-PR to forecast the occurrence of osteoporosis within the training and validation cohorts were 90.8% [95% confidence interval (CI), 88.4-93.2%] vs. 78.0% (95% CI, 76.0-79.9%) and 94.6% (95% CI, 89.3-99.8%) vs. 88.8% (95% CI, 86.2-91.5%), respectively.

Conclusions

The osteoporosis prediction model using autosegmentation of proximal femur and machine-learning analysis with PyRadiomics features on APCT showed excellent diagnostic feasibility and technical success.

Evaluation and analysis of surgical treatment for single-level or multi-level lumbar degenerative disease based on radiography

Background

Radiography has a low level of radiation exposure while providing valuable information. Due to its cost effectiveness and widespread availability, the preoperative radiographic imaging examination is a valuable approach for assessing patients with spinal disease. This study aimed to examine the influence of preoperative X-ray evaluation on the surgical treatment of patients with single- or multi-level lumbar degenerative disease (LDD).

Methods

A retrospective cohort analysis was conducted of 172 patients diagnosed with LDD who underwent transforaminal lumbar interbody fusion (TLIF) or posterior lumbar interbody fusion (PLIF) surgery between December 2021 and February 2023 at the Shanghai Changzheng Hospital. Various parameters were measured on preoperative radiographs, including the iliac crest height, median iliac angle (MIA), lumbar lordosis (LL), intervertebral facet joint degeneration, lumbosacral angle (LSA), intervertebral foramen height (IFH), and surgical segment. The surgical treatment was evaluated based on the operative time, intraoperative blood loss, and postoperative complications. A correlation analysis and independent sample t-tests were used to assess the relationship between preoperative radiographic variables and surgical treatments. Further, a multivariate linear regression analysis was employed to identify the risk factors affecting the clinical outcomes.

Results

The correlation analysis and t-test results showed that the MIA, height of the iliac crest, intervertebral facet joint degeneration, and surgical segment were significantly correlated with the surgical treatments (P<0.05). Specifically, the height of the iliac crest, intervertebral facet joint degeneration, and surgical segment were positively correlated with the surgical treatments. Conversely, the MIA was negatively correlated with the surgical treatments. However, no significant differences were observed between the IFH, LSA, and LL in relation to posterior lumbar surgery (P>0.05). The multiple linear regression analysis showed that the height of the iliac crest, MIA, intervertebral facet joint degeneration, and surgical segment were independent factors affecting the surgical treatments of patients with single- or multi-level LDD. These findings highlight the importance of considering these factors when planning and performing lumbar surgery.

Conclusions

The measurements taken from radiographs, including the height of the iliac crest, MIA, intervertebral facet joint degeneration, and surgical segment, demonstrate potential influences on the treatment of single- and multi-level lumbar spine surgery. These variables can be captured in plain film imaging and can provide valuable insights into the surgical procedure and offer guidance for the operation. By analyzing these radiographic measurements, surgeons can gain a better understanding of a patient's condition and tailor the surgical approach accordingly, thus optimizing the outcomes of the surgery.

Accuracy of an artificial intelligence algorithm for detecting moderate-to-severe vertebral compression fractures on abdominal and thoracic computed tomography scans

Objective

To describe the accuracy of HealthVCF, a software product that uses artificial intelligence, in the detection of incidental moderate-to-severe vertebral compression fractures (VCFs) on chest and abdominal computed tomography scans.

Materials and Methods

We included a consecutive sample of 899 chest and abdominal computed tomography scans of patients 51-99 years of age. Scans were retrospectively evaluated by the software and by two specialists in musculoskeletal imaging for the presence of VCFs with vertebral body height loss > 25%. We compared the software analysis with that of a general radiologist, using the evaluation of the two specialists as the reference.

Results

The software showed a diagnostic accuracy of 89.6% (95% CI: 87.4-91.5%) for moderate-to-severe VCFs, with a sensitivity of 73.8%, a specificity of 92.7%, and a negative predictive value of 94.8%. Among the 145 positive scans detected by the software, the general radiologist failed to report the fractures in 62 (42.8%), and the algorithm detected additional fractures in 38 of those scans.

Conclusion

The software has good accuracy for the detection of moderate-to-severe VCFs, with high specificity, and can increase the opportunistic detection rate of VCFs by radiologists who do not specialize in musculoskeletal imaging.

Vertebral Fracture Risk Thresholds from Phantom-Less Quantitative Computed Tomography-Based Finite Element Modeling Correlate to Phantom-Based Outcomes

INTRODUCTION

Osteoporosis indicates weakened bones and heightened fracture susceptibility due to diminished bone quality. Dual-energy x-ray absorptiometry is unable to assess bone strength. Volumetric bone mineral density (vBMD) from quantitative computed tomography (QCT) has been used to establish guidelines as equivalent measurements for osteoporosis. QCT-based finite element analysis (FEA) has been implemented using calibration phantoms to establish bone strength thresholds based on the established vBMD. The primary aim was to validate vertebral failure load thresholds using a phantom-less approach with previously established thresholds, advancing a phantom-free approach for fracture risk prediction.

METHODOLOGY

A controlled cohort of 108 subjects (68 females) was used to validate sex-specific vertebral fracture load thresholds for normal, osteopenic, and osteoporotic subjects, obtained using a QCT/FEA-based phantom-less calibration approach and two material equations.

RESULTS

There were strong prediction correlations between the phantom-less and phantom-based methods (R2: 0.95 and 0.97 for males, and R2: 0.96 and 0.98 for females) based on the two equations. Bland Altman plots and paired t-tests showed no significant differences between methods. Predictions for bone strengths and thresholds using the phantom-less method matched those obtained using the phantom calibration and those previously established, with ≤4500 N (fragile) and ≥6000 N (normal) bone strength in females, and ≤6500 N (fragile) and ≥8500 N (normal) bone strength in males.

CONCLUSION

Phantom-less QCT-based FEA can allow for prospective and retrospective studies evaluating incidental vertebral fracture risk along the spine and their association with spine curvature and/or fracture etiology. The findings of this study further supported the application of phantom-less QCT-based FEA modeling to predict vertebral strength, aiding in identifying individuals prone to fractures. This reinforces the rationale for adopting this method as a comprehensive approach in predicting and managing fracture risk.

Towards Improved Identification of Vertebral Fractures in Routine Computed Tomography (CT) Scans: Development and External Validation of a Machine Learning Algorithm

Vertebral fractures (VFs) are the hallmark of osteoporosis, being one of the most frequent types of fragility fracture and an early sign of the disease. They are associated with significant morbidity and mortality. VFs are incidentally found in one out of five imaging studies, however, more than half of the VFs are not identified nor reported in patient computed tomography (CT) scans. Our study aimed to develop a machine learning algorithm to identify VFs in abdominal/chest CT scans and evaluate its performance. We acquired two independent data sets of routine abdominal/chest CT scans of patients aged 50 years or older: a training set of 1011 scans from a non-interventional, prospective proof-of-concept study at the Universitair Ziekenhuis (UZ) Brussel and a validation set of 2000 subjects from an observational cohort study at the Hospital of Holbaek. Both data sets were externally reevaluated to identify reference standard VF readings using the Genant semiquantitative (SQ) grading. Four independent models have been trained in a cross-validation experiment using the training set and an ensemble of four models has been applied to the external validation set. The validation set contained 15.3% scans with one or more VF (SQ2-3), whereas 663 of 24,930 evaluable vertebrae (2.7%) were fractured (SQ2-3) as per reference standard readings. Comparison of the ensemble model with the reference standard readings in identifying subjects with one or more moderate or severe VF resulted in an area under the receiver operating characteristic curve (AUROC) of 0.88 (95% confidence interval [CI], 0.85-0.90), accuracy of 0.92 (95% CI, 0.91-0.93), kappa of 0.72 (95% CI, 0.67-0.76), sensitivity of 0.81 (95% CI, 0.76-0.85), and specificity of 0.95 (95% CI, 0.93-0.96). We demonstrated that a machine learning algorithm trained for VF detection achieved strong performance on an external validation set. It has the potential to support healthcare professionals with the early identification of VFs and prevention of future fragility fractures. © 2023 UCB S.A. and The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Bone mineral density estimation from a plain X-ray image by learning decomposition into projections of bone-segmented computed tomography

Osteoporosis is a prevalent bone disease that causes fractures in fragile bones, leading to a decline in daily living activities. Dual-energy X-ray absorptiometry (DXA) and quantitative computed tomography (QCT) are highly accurate for diagnosing osteoporosis; however, these modalities require special equipment and scan protocols. To frequently monitor bone health, low-cost, low-dose, and ubiquitously available diagnostic methods are highly anticipated. In this study, we aim to perform bone mineral density (BMD) estimation from a plain X-ray image for opportunistic screening, which is potentially useful for early diagnosis. Existing methods have used multi-stage approaches consisting of extraction of the region of interest and simple regression to estimate BMD, which require a large amount of training data. Therefore, we propose an efficient method that learns decomposition into projections of bone-segmented QCT for BMD estimation under limited datasets. The proposed method achieved high accuracy in BMD estimation, where Pearson correlation coefficients of 0.880 and 0.920 were observed for DXA-measured BMD and QCT-measured BMD estimation tasks, respectively, and the root mean square of the coefficient of variation values were 3.27 to 3.79% for four measurements with different poses. Furthermore, we conducted extensive validation experiments, including multi-pose, uncalibrated-CT, and compression experiments toward actual application in routine clinical practice.

What is the role of CT-based Hounsfield unit assessment in the evaluation of bone mineral density in patients undergoing 1- or 2-level lumbar spinal fusion for degenerative spinal pathologies? A prospective study

BACKGROUND CONTEXT

Computed tomography-based vertebral attenuation values (CT-based HU) have been shown to correlate with T-scores on DEXA scan; and have been acknowledged as an independent factor for predicting fragility fractures. Most patients undergoing lumbar surgeries require CT as part of their preoperative evaluation.

PURPOSE

The current study was thus planned to evaluate the role of lumbar CT as an opportunistic investigation in determining BMD preoperatively in patients undergoing lumbar fusion.

STUDY DESIGN

Prospective cohort study.

PATIENT SAMPLE

Patients older than 45 years, who underwent one- to two-level lumbar (L3-S1 levels) fusions.

OUTCOME MEASURES

Comparison of the quantitative assessment of osteoporosis using Hounsfield Units (HU) on CT (L1-L5) and mean lumbar T-scores on DEXA (Dual Energy X-ray Absorptiometry).

HYPOTHESIS

HU on CT is comparable to T-score on DEXA as a suitable modality for the assessment of osteoporosis in patients undergoing one- to two-level lumbar fusion.

METHODS

A prospective cohort study was conducted between January and December 2021. Patients older than 45 years, who underwent one- to two-level lumbar (L3-S1 levels) fusions and had complete clinico-radiological records, were prospectively enrolled. A comparison was drawn between the HU (measured by placing an oval region of interest [ROI] over axial, sagittal and coronal images of lumbar vertebrae) on CT and T-scores on DEXA, and analyzed statistically. The HU values correlating best with normal (group A), osteopenia (B) and osteoporosis (C) categories (classified based on T-scores of lumbar spines) were determined statistically.

RESULTS

Overall, 87 patients (mean age of 60.56±11.63 years; 63 [72.4%] female patients) were prospectively studied. There was a statistically significant difference in the mean age (p=.01) and sex distribution (predominantly female patients; p=.03) of patients belonging to groups B (osteopenic) and C (osteoporotic patients), as compared with group A. The greatest correlation between T-score (on DEXA) and HU (on CT) for differentiating osteopenia (group B) from group A was observed at levels L1 (p<.001), L2 (p<.001) and L3 (p<.001). Based on receiver-operating characteristic (ROC) curve analysis, the cut-off values for HU for identifying osteopenia were 159 (at L1; sensitivity 81.6 and specificity 80) and 162 (at L2; sensitivity 80 and specificity 71.1). In addition, there was statistically significant correlation between T-score (on DEXA) and HU at all the lumbar levels for distinguishing osteoporosis (group C), although the difference was most evident at the upper lumbar (L1 and L2) levels (p<.001). Based on ROC analysis, cut-off HU values for defining osteoporosis were 127 (at L1; sensitivity 71.3 and specificity 70) and 117 (at L2; sensitivity 65.5 and specificity 90).

CONCLUSION

Based on our study, the measurement of HU on CT at upper lumbar levels can be considered as "surrogate marker" for BMD in the diagnosis of osteopenia (cut-off: 159 at L1, 162 at L2) and osteoporosis (cut-off: 127 at L1, 117 at L2) in patients undergoing lumbar fusion surgeries. The HU measurements on CT at the lower lumbar levels (L4 and L5) are less reliable in this preoperative scenario.

Addressing Challenges of Opportunistic Computed Tomography Bone Mineral Density Analysis

Computed tomography (CT) offers advanced biomedical imaging of the body and is broadly utilized for clinical diagnosis. Traditionally, clinical CT scans have not been used for volumetric bone mineral density (vBMD) assessment; however, computational advances can now leverage clinically obtained CT data for the secondary analysis of bone, known as opportunistic CT analysis. Initial applications focused on using clinically acquired CT scans for secondary osteoporosis screening, but opportunistic CT analysis can also be applied to answer research questions related to vBMD changes in response to various disease states. There are several considerations for opportunistic CT analysis, including scan acquisition, contrast enhancement, the internal calibration technique, and bone segmentation, but there remains no consensus on applying these methods. These factors may influence vBMD measures and therefore the robustness of the opportunistic CT analysis. Further research and standardization efforts are needed to establish a consensus and optimize the application of opportunistic CT analysis for accurate and reliable assessment of vBMD in clinical and research settings. This review summarizes the current state of opportunistic CT analysis, highlighting its potential and addressing the associated challenges.

Opportunistic AI-enabled automated bone mineral density measurements in lung cancer screening and coronary calcium scoring CT scans are equivalent

Rationale and objectives

We previously reported a novel manual method for measuring bone mineral density (BMD) in coronary artery calcium (CAC) scans and validated our method against Dual X-Ray Absorptiometry (DEXA). Furthermore, we have developed and validated an artificial intelligence (AI) based automated BMD (AutoBMD) measurement as an opportunistic add-on to CAC scans that recently received FDA approval. In this report, we present evidence of equivalency between AutoBMD measurements in cardiac vs lung CT scans.

Materials and methods

AI models were trained using 132 cases with 7649 (3 mm) slices for CAC, and 37 cases with 21918 (0.5 mm) slices for lung scans. To validate AutoBMD against manual measurements, we used 6776 cases of BMD measured manually on CAC scans in the Multi-Ethnic Study of Atherosclerosis (MESA). We then used 165 additional cases from Harbor UCLA Lundquist Institute to compare AutoBMD in patients who underwent both cardiac and lung scans on the same day.

Results

Mean±SD for age was 69 ± 9.4 years with 52.4% male. AutoBMD in lung and cardiac scans, and manual BMD in cardiac scans were 153.7 ± 43.9, 155.1 ± 44.4, and 163.6 ± 45.3 g/cm3, respectively (p = 0.09). Bland-Altman agreement analysis between AutoBMD lung and cardiac scans resulted in 1.37 g/cm3 mean differences. Pearson correlation coefficient between lung and cardiac AutoBMD was R2 = 0.95 (p < 0.0001).

Conclusion

Opportunistic BMD measurement using AutoBMD in CAC and lung cancer screening scans is promising and yields similar results. No extra radiation plus the high prevalence of asymptomatic osteoporosis makes AutoBMD an ideal screening tool for osteopenia and osteoporosis in CT scans done for other reasons.

Combination of vertebral bone quality scores from different magnetic resonance imaging sequences improves prognostic value for the estimation of osteoporosis

BACKGROUND CONTEXT

Recent findings revealed a correlation between vertebral bone quality based on T1-weighted (VBQ_T1) magnetic resonance imaging (MRI) and volumetric bone mass density (vBMD) measured using quantitative computerized tomography. The coherence of VBQ for other MRI sequences, such as T2 or short tau inversion recovery (STIR), has not been examined. The combination of different VBQs has not been studied.

PURPOSE

The aims of the study were to confirm the correlation between VBQ_T1 and vBMD and to examine VBQs from other MRI sequences and their combination with vBMD.

STUDY DESIGN/SETTING

This was a retrospective cross-sectional study.

PATIENT SAMPLE

The sample consisted of patients older than 18 years, who received treatment at a level-one university spine center of the German Spine Society for degenerative or traumatic reasons in 2017-2021.

OUTCOME MEASURES

The outcome measures were the correlation of VBQs from different MRI sequences with vBMD and the association of VBQs with osteopenia/osteoporosis.

METHODS

Patients' VBQ was calculated based on the signal intensities of the vertebral bodies L1-4 in T1-, T2-, and STIR-weighted MRI. The VBQ was standardized according to the signal intensity of the cerebrospinal fluid. The vBMD was determined using data from a calibrated scanner (SOMATOM Definition AS+) and processed with CliniQCT (Mindways Software, Inc., USA). Groups were divided according to vBMD into the following groups: (I) osteoporosis/osteopenia (< 120 mg/m³) and (II) healthy (≥120 mg/m³). An analysis of the correlation between various VBQs and vBMD as well as receiver operating characteristic (ROC) and binary regression analyses were performed for the prediction of osteoporosis/osteopenia.

RESULTS

We included 136 patients (women: 56.6%) in the study (69.7 ± 15.0 years). According to vBMD, 108 patients (79.4%) had osteoporosis/osteopenia. Women were affected significantly more often than men (p = .045) and had significantly higher VBQ_T1 and VBQ_T2 values than men (VBQ_T1: p = .048; VBQ_T2: p = .013). VBQ_T1 and VBQ_T2 values were significantly higher in patients with osteoporosis/osteopenia than in healthy persons (VBQ_T1: p<.001; VBQ_T2: p = .025). VBQ_T1 and VBQ_T2 were significantly negatively correlated with vBMD with a moderate effect size (p<.001), while VBQ_STIR was not significantly correlated with vBMD, although it showed a positive coherence. The combination of different VBQs in terms of VBQ_T1 × VBQ_T2 / VBQ_STIR distinctly increased the effect size of the negative correlation with vBMD compared to VBQ alone. A cutoff value for VBQ_T1 × VBQ_T2 / VBQ_STIR of 2.9179 achieved a sensitivity of 80.0%, a specificity of 75.0%, and an area under the curve (AUC) of 0.775 for the determination of osteoporosis. The mathematical model derived from the binary logistic regression showed an excellent AUC of 0.846.

CONCLUSIONS

This study confirms a significant correlation between VBQ_T1 and vBMD. The combination of VBQs from different MRI sequences enhances the prognostic value of VBQ for the determination of osteoporosis. While safe clinical application of VBQ for the determination of osteoporosis requires further validation, VBQ might offer opportunistic estimation for further diagnostics.

Improvement of Patient Classification Using Feature Selection Applied to Bidirectional Axial Transmission

Osteoporosis is still a worldwide problem, particularly due to associated fragility fractures. Patients at risk of fracture are currently detected using the X-Ray gold standard dual-energy X-ray absorptiometry (DXA), based on a calibrated 2-D image. Different alternatives, such as 3-D X-rays, magnetic resonance imaging (MRI) or ultrasound, have been proposed, the latter having advantages of being portable and sensitive to mechanical and geometrical properties. Bidirectional axial transmission (BDAT) has been used to classify between patients with or without nontraumatic fractures using "classical" ultrasonic parameters, such as velocities, as well as cortical thickness and porosity, obtained from an inverse problems. Recently, complementary parameters acquired with structural and textural analysis of guided wave spectrum images (GWSIs) have been introduced. These parameters are not limited by solution ambiguities, as for inverse problem. The aim of the study is to improve the patient classification using a feature selection strategy for all available ultrasound features completed by clinical parameters. To this end, three classical feature ranking methods were considered: analysis of variance (ANOVA), recursive feature elimination (RFE), and extreme gradient boosting importance feature (XGBI). In order to evaluate the performance of the feature selection techniques, three classical classification methods were used: logistic regression (LR), support vector machine (SVM), and extreme gradient boosting (XGB). The database was obtained from a previous clinical study [Minonzio et al., 2019]. Results indicate that the best accuracy of 71 [66-76]% was achieved by using RFE and SVM with 22 (out of 43) ultrasonic and clinical features. This value outperformed the accuracy of 68 [64-73]% reached with 2 (out of 6) DXA and clinical features. These values open promising perspectives toward improved and generalizable classification of patients at risk of fracture.

How to implement guidelines and models of care

In subjects older than 50 years, the presence of clinical risk factors (CRFs) for fractures or a recent fracture is the cornerstone for case finding. In patients who are clinically at high short- and long-term risk of fractures (those with a recent clinical fracture or with multiple CRFs), further assessment with bone mineral density (BMD) measurement using dual-energy absorptiometry (DXA), imaging of the spine, fall risk evaluation and laboratory examination contributes to treatment decisions according to the height and modifiability of fracture risk. Treatment is available with anti-resorptive and anabolic drugs, and from the start of treatment a lifelong strategy is needed to decide about continuous, intermittent, and sequential therapy. Implementation of guidelines requires further initiatives for improving case finding, public awareness about osteoporosis and national policies on reimbursement of assessment and therapy.

Definition of normal vertebral morphometry using NHANES‐II radiographs

A robust definition of normal vertebral morphometry is required to confidently identify abnormalities such as fractures. The Second National Health and Nutrition Examination Survey (NHANES‐II) collected a nationwide probability sample to document the health status of the United States. Over 10,000 lateral cervical spine and 7,000 lateral lumbar spine X‐rays were collected. Demographic, anthropometric, health, and medical history data were also collected. The coordinates of the vertebral body corners were obtained for each lumbar and cervical vertebra using previously validated, automated technology consisting of a pipeline of neural networks and coded logic. These landmarks were used to calculate six vertebral body morphometry metrics. Descriptive statistics were generated and used to identify and trim outliers from the data. Descriptive statistics were tabulated using the trimmed data for use in quantifying deviation from average for each metric. The dependency of these metrics on sex, age, race, nation of origin, height, weight, and body mass index (BMI) was also assessed. There was low variation in vertebral morphometry after accounting for vertebrae (eg, L1, L2), and the R² was high for ANOVAs. Excluding outliers, age, sex, race, nation of origin, height, weight, and BMI were statistically significant for most of the variables, though the F‐statistic was very small compared to that for vertebral level. Excluding all variables except vertebra changed the ANOVA R² very little. Reference data were generated that could be used to produce standardized metrics in units of SD from mean. This allows for easy identification of abnormalities resulting from vertebral fractures, atypical vertebral body morphometries, and other congenital or degenerative conditions. Standardized metrics also remove the effect of vertebral level, facilitating easy interpretation and enabling data for all vertebrae to be pooled in research studies. © 2022 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Imaging of Metabolic Bone Diseases: The Spine View, Part II

Metabolic bone diseases comprise a wide spectrum. Osteoporosis, the most frequent, characteristically involves the spine, with a high impact on health care systems and on the morbidity of patients due to the occurrence of vertebral fractures (VFs).Part II of this review completes an overview of state-of-the-art techniques on the imaging of metabolic bone diseases of the spine, focusing on specific populations and future perspectives. We address the relevance of diagnosis and current status on VF assessment and quantification. We also analyze the diagnostic techniques in the pediatric population and then review the assessment of body composition around the spine and its potential application. We conclude with a discussion of the future of osteoporosis screening, through opportunistic diagnosis and the application of artificial intelligence.

A software program for automated compressive vertebral fracture detection on elderly women's lateral chest radiograph: Ofeye 1.0

Background

Because osteoporotic vertebral fracture (OVF) on chest radiographs is commonly missed in radiological reports, we aimed to develop a software program which offers automated detection of compressive vertebral fracture (CVF) on lateral chest radiographs, and which emphasizes CVF detection specificity with a low false positivity rate.

Methods

For model training, we retrieved 3,991 spine radiograph cases and 1,979 chest radiograph cases from 16 sources, with among them in total 1,404 cases had OVF. For model testing, we retrieved 542 chest radiograph cases and 162 spine radiograph cases from four independent clinics, with among them 215 cases had OVF. All cases were female subjects, and except for 31 training data cases which were spine trauma cases, all the remaining cases were post-menopausal women. Image data included DICOM (Digital Imaging and Communications in Medicine) format, hard film scanned PNG (Portable Network Graphics) format, DICOM exported PNG format, and PACS (Picture Archiving and Communication System) downloaded resolution reduced DICOM format. OVF classification included: minimal and mild grades with <20% or ≥20-25% vertebral height loss respectively, moderate grade with ≥25-40% vertebral height loss, severe grade with ≥40%-2/3 vertebral height loss, and collapsed grade with ≥2/3 vertebral height loss. The CVF detection base model was mainly composed of convolution layers that include convolution kernels of different sizes, pooling layers, up-sampling layers, feature merging layers, and residual modules. When the model loss function could not be further decreased with additional training, the model was considered to be optimal and termed 'base-model 1.0'. A user-friendly interface was also developed, with the synthesized software termed 'Ofeye 1.0'.

Results

Counting cases and with minimal and mild OVFs included, base-model 1.0 demonstrated a specificity of 97.1%, a sensitivity of 86%, and an accuracy of 93.9% for the 704 testing cases. In total, 33 OVFs in 30 cases had a false negative reading, which constituted a false negative rate of 14.0% (30/215) by counting all OVF cases. Eighteen OVFs in 15 cases had OVFs of ≥ moderate grades missed, which constituted a false negative rate of 7.0% (15/215, i.e., sensitivity 93%) if only counting cases with ≥ moderate grade OVFs missed. False positive reading was recorded in 13 vertebrae in 13 cases (one vertebra in each case), which constituted a false positivity rate of 2.7% (13/489). These vertebrae with false positivity labeling could be readily differentiated from a true OVF by a human reader. The software Ofeye 1.0 allows 'batch processing', for example, 100 radiographs can be processed in a single operation. This software can be integrated into hospital PACS, or installed in a standalone personal computer.

Conclusions

A user-friendly software program was developed for CVF detection on elderly women's lateral chest radiographs. It has an overall low false positivity rate, and for moderate and severe CVFs an acceptably low false negativity rate. The integration of this software into radiological practice is expected to improve osteoporosis management for elderly women.

Understanding current UK practice for the incidental identification of vertebral fragility fractures from CT scans: an expert elicitation study

Background

There is an emerging interest in using automated approaches to enable the incidental identification of vertebral fragility fractures (VFFs) on existing medical images visualising the spine.

Aim

To quantify values, and the degree of uncertainty associated with them, for the incidental identification of VFFs from computed tomography (CT) scans in current practice.

Methods

An expert elicitation exercise was conducted to generate point estimates and measures of uncertainty for four values representing the probability of: VFF being correctly reported by the radiologist; the absence of VFF being correctly assessed by the radiologist; being referred for management when a VFF is identified; having a dual-energy X-ray absorptiometry (DXA) scan after general practitioner (GP) referral. Data from a sample of seven experts in the diagnosis and management of people with VFFs were pooled using mathematical aggregation.

Results

The estimated mean values for each probability parameter were: VFF being correctly reported by the radiologist = 0.25 (standard deviation (SD): 0.21); absence of VFF being correctly assessed by the radiologist = 0.89 (0.10); being referred for management when a VFF is identified by the radiologist = 0.15 (0.12); having a DXA scan after GP referral = 0.66 (0.28).

Discussion

These estimates could be used to facilitate the subsequent early economic evaluation of potential new approaches to improve the health outcomes of people with VFFs.

Conclusion

In the absence of epidemiological studies, this study produced point estimates and measures of uncertainty for key parameters needed to describe current pathways for the incidental diagnosis of VFFs.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40520-022-02124-w.

BMD accuracy errors specific to phantomless calibration of CT scans of the lumbar spine

Opportunistic screening using existing CT images may be a new strategy to identify subjects at increased risk for osteoporotic fracture. Low bone mineral density (BMD) is a key parameter but routine clinical CT scans do not include a calibration phantom to calculate BMD from the measured CT values. An alternative is internal or phantomless calibration, which is based on the CT values of air and of internal tissues of the subject such as blood, muscle or adipose tissue. However, the composition and as a consequence the CT values of these so-called internal calibration materials vary among subjects, which introduces additional BMD accuracy errors compared to phantom based calibration. The objective of this study was to quantify these accuracy errors and to identify optimum combinations of internal calibration materials (IM) for BMD assessments in opportunistic screening. Based on the base material decomposition theory we demonstrate how BMD can be derived from the CT values of the internal calibration materials. 121 CT datasets of the lumbar spine form postmenopausal women were used to determine the population variance of blood assessed in the aorta or the inferior vena cava, skeletal muscle of the erector spinae or psoas, subcutaneous adipose tissue (SAT) and air. The corresponding standard deviations were used for error propagation to determine phantomless calibration related BMD accuracy errors. Using a CT value of 150 HU, a typical value of trabecular bone, simulated BMD accuracy errors for most IM combinations containing air as one of the two base materials were below 5% or 6 mg/cm³. The lowest errors were determined for the combination of blood and air (<2 mg/cm³). The combination of blood and skeletal muscle resulted in higher errors (>10.5% or >12 mg/cm³) and is not recommended. Due to possible age-related differences in tissue composition, the selection of IMs is suggested to be adapted according to the measured subject. In younger subjects without significant aortic calcifications, air and blood of the aorta may be the best combination whereas in elderly subjects, air and SAT (error of 4%) may be preferable. The use of skeletal muscle as one of the two IMs is discouraged, in particular in elderly subjects because of varying fatty infiltration. A practical implementation of the internal calibration with different IM pairs confirmed the theoretical results. In summary, compared to a phantom based calibration the phantomless approach used for opportunistic screening creates additional BMD accuracy errors of 2% or more, dependent on the used internal reference tissues. The impact on fracture prediction still must be evaluated.

Fully automated opportunistic screening of vertebral fractures and osteoporosis on more than 150,000 routine computed tomography scans

OBJECTIVE

Osteoporosis is underdiagnosed and undertreated, although severe complications of osteoporotic fractures, including vertebral fractures, are well known. This study sought to assess the feasibility and results of an opportunistic screening of vertebral fractures and osteoporosis in a large database of lumbar or abdominal CT scans.

MATERIAL AND METHODS

Data were analyzed from CT scans obtained in 35 hospitals from patients aged 60 years and more and stored in a Picture Archiving and Communication System in Assistance-Publique-Hôpitaux de Paris, from 2007 to 2013. Dedicated software analyzed the presence of at least 1 vertebral fracture (VF), and measured Hounsfield Units (HU) in lumbar vertebrae. A simulated T-score was calculated.

RESULTS

Data were analyzed from 152 268 patients (73.2 ± 9.07 years). Success rates for VF assessment and HU measurements were 82 and 87% respectively. Prevalence of VF was 24.5% and increased with age. Areas under the receiver operating characteristic curves for the detection of VF were 0.61 and 0.62 for mean HU of lumbar vertebrae and L1 HU, respectively. In patients without VF, HU decreased with age, similarly in males and females. The prevalence of osteoporosis (sT-score ≤ - 2.5) was 23.8% and 36.5% in patients without and with VFs respectively.

CONCLUSION

Opportunistic screening in patients 60 years and older having lumbar or abdominal CT scans is feasible at large scale to screen vertebral fractures and osteoporosis.