Routine MRI for opportunistic osteoporosis screening: diagnostic accuracy of vertebral bone quality and a novel femoral bone quality score compared with DXA.
Aug 2026· Skeletal Radiology· 0 citations· 18 references
Medicine
TL;DR
VBQ can serve as an opportunistic exploratory screening marker of low BMD and FBQcsf may complement VBQ by identifying early marrow-related changes, although further validation is required.
Abstract
Objectives
To evaluate vertebral bone quality (VBQ) and femoral bone quality (FBQ) scores as potential imaging markers of bone mineral density (BMD) for opportunistic osteoporosis screening in at‑risk patients and to assess their correlation with DXA T-scores.
Methods
This retrospective cross-sectional study included women ≥ 40 years and men ≥ 50 years who underwent DXA and lumbar spine, hip, or pelvic MRI within 60 days. All MRI examinations were performed at 1.5 T. VBQ and FBQ were obtained using manually placed circular ROIs. FBQ was derived from coronal T1-weighted images using two approaches: femoroacetabular joint fluid (FBQfa) and cerebrospinal fluid (FBQcsf). Interobserver reproducibility was assessed using intraclass correlation coefficient (ICC). Correlations with DXA T-scores and diagnostic performance for osteoporosis were analyzed using Pearson coefficients and ROC curves.
Results
Among 248 patients, VBQ demonstrated excellent reproducibility (ICC = 0.922) and a weak-to-moderate negative correlation with DXA T-scores (r = -0.307; p < 0.001). VBQ values increased across DXA bone density categories, with an optimal cutoff of 3.61 for osteoporosis (AUC = 0.71). FBQcsf showed high reproducibility (ICC = 0.874) compared with FBQfa (ICC = 0.462) but no evidence of correlation with DXA (p = 0.151). FBQcsf distinguished normal BMD from osteopenia (p = 0.002) and, after adjustment for age and sex, also differentiated patients with osteopenia from those with osteoporosis (p = 0.011).
Conclusion
VBQ can serve as an opportunistic exploratory screening marker of low BMD. FBQcsf may complement VBQ by identifying early marrow-related changes, although further validation is required.
OBJECTIVES
Accurate detection of impaired vertebral bone quality (VBQ) is essential for identifying patients at risk of osteoporosis-related fractures and for optimising surgical planning. Dual-energy X-ray absorptiometry (DXA) is the gold standard but is limited by degenerative artefacts and restricted availability. CT-derived Hounsfield units (HU) and MRI-derived VBQ scores provide alternatives for assessing bone quality. This meta-analysis compared HU and VBQ performance for identifying osteopenia and osteoporosis.
MATERIALS AND METHODS
PubMed, EMBASE, MEDLINE, and Cochrane Library were searched to September 2025. Studies reporting HU or VBQ referenced to DXA or quantitative CT were included. Continuous outcomes were pooled as standardised mean differences (SMD), and diagnostic accuracy was evaluated using bivariate hierarchical summary receiver operating characteristic models, threshold synthesis, meta-regression, and bootstrap validation.
RESULTS
Forty-two studies (9214 participants) were included. Both HU and VBQ distinguished impaired from normal bone quality, although HU demonstrated larger SMDs (osteopenia/osteoporosis -2.62; osteoporosis -1.53) than VBQ (0.77 and 0.81). For osteopenia, pooled areas under the ROC curve (AUCs) were 0.87 for HU and 0.75 for VBQ; for osteoporosis, AUCs were 0.86 and 0.79, respectively. Threshold synthesis identified pooled cut-offs for both modalities: 138.9 HU (osteopenia), 104.6 HU (osteoporosis), and VBQ thresholds of 2.83 and 3.14. Validation analyses consistently favoured HU, while VBQ demonstrated greater variability.
CONCLUSION
Both HU and VBQ offer useful assessments of vertebral bone quality, but HU provides higher and more stable diagnostic accuracy. HU thresholds may support opportunistic osteoporosis screening, but their clinical applicability remains uncertain given inter-study heterogeneity, while further refinement of MRI-based VBQ methods is warranted.
KEY POINTS
Question How do computed tomography attenuation and magnetic resonance imaging vertebral bone quality scoring compare in identifying osteopenia and osteoporosis on routine imaging? Findings Computed tomography attenuation measurements outperformed magnetic resonance imaging bone quality scoring, demonstrating higher sensitivity, specificity, and accuracy for detecting osteopenia and osteoporosis. Clinical relevance Routine CT and MRI can support opportunistic detection of low bone density, enabling earlier intervention, improved fracture risk stratification, and more informed preoperative assessment without additional imaging or radiation.
Omar Lubbad, Akram Hagos, Laila Lubbad et al.· European Radiology· 0 citations
OBJECTIVE
MRI-derived vertebral bone quality (VBQ) scores estimate vertebral marrow-related bone quality on T1-weighted imaging and may complement dual-energy X-ray absorptiometry (DXA) in spine patients. Discordance between VBQ and DXA is common, particularly in the lumbar spine, where degenerative changes may bias DXA measurements. This study aimed to identify physiological determinants associated with lumbar VBQ-DXA discordance.
METHODS
We retrospectively included 266 consecutive adults with preoperative lumbar MRI and DXA performed within 6 months before surgery. Lumbar VBQ was calculated as the ratio of the median T1 signal intensity of the L1-L4 vertebral bodies to cerebrospinal fluid signal intensity. Correlations between VBQ, DXA-derived T-scores, and clinical and laboratory variables were examined, and multivariable regression was used to identify independent determinants of VBQ. Concordance and discordance phenotypes were defined using the lower T-score from the lumbar spine and femoral neck (DXA-normal: ≥ -1.0; DXA-abnormal: < -1.0) and an ROC-derived VBQ threshold.
RESULTS
VBQ was positively correlated with age (r = 0.706, p < 0.001) and moderately inversely correlated with both lumbar spine T-score (r = -0.396, p < 0.001) and femoral neck T-score (r = -0.374, p < 0.001). ROC analysis showed moderate discrimination of DXA-defined osteoporosis, with an AUC of 0.728. At the cutoff of 2.685, sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy were 65.1%, 70.8%, 30.1%, 91.3%, and 69.9%, respectively. In multivariable regression, triglycerides (β = 0.397, p = 0.003), very low-density lipoprotein cholesterol (β = 1.325, p = 0.002), glycated albumin (β = 0.184, p = 0.003), and conjugated bilirubin (β = 0.162, p = 0.004) were independent determinants of VBQ. The DXA-normal/VBQ-abnormal phenotype showed enrichment of this TG/VLDL-C/GA/CB metabolic pattern.
CONCLUSIONS
Lumbar VBQ-DXA discordance is associated with a physiological signature characterized by triglyceride-rich lipoproteins, glycation burden, and conjugated bilirubin. VBQ may complement DXA by identifying metabolically vulnerable, marrow-related bone quality changes that are not fully captured by DXA-derived T-scores.
Jie Hao, Fayao Yan, Yu Yao et al.· Bone· 0 citations
Introduction CT-based Hounsfield units (HU) and MRI-based vertebral bone quality (VBQ)-scores are opportunistic markers of bone mineral density and potential alternatives for dual-energy X-ray absorptiometry (DEXA). Research question What is the correlation between HU values and VBQ-scores in relation to DEXA, and what is the corresponding area-under-the-curve (AUC)-based diagnostic discrimination to detect osteopenia and osteoporosis? Material and methods PubMed, Embase, Web of Science, Cochrane Library, and CINAHL were searched through August 2025 for studies reporting correlations between DEXA and CT and/or MRI or the AUC-based diagnostic discrimination of CT and/or MRI to detect osteopenia and osteoporosis. Random effects models were used to calculate pooled correlation coefficients and AUC values along with 95% confidence intervals. Results Forty-three studies, including 8222 patients, were analyzed. HU demonstrated stronger correlations with DEXA T-scores than the VBQ-score (maximum correlation coefficient: 0.741 (95% CI: 0.578-0.847) vs. −0.476 (95% CI: −0.548 to −0.398) and greater AUC-based diagnostic discrimination for osteopenia and osteoporosis (maximum AUC: 0.875 (95% CI: 0.841-0.909) vs. 0.806 (95% CI: 0.684-0.929) across all vertebrae and DEXA sites. Correlation and AUC-based diagnostic discrimination of HU were consistent across vertebrae, with slightly better results at L2 and the mean of L1–L4. VBQ-score correlations were highest at L1 and gradually decreased from L1 to L4. Discussion and conclusion When CT is available, HU appears to be the better-supported opportunistic marker for low BMD than VBQ. All vertebrae demonstrated comparable correlation and AUC-based diagnostic discrimination, with L2 HU and the mean of L1-L4 HU showing slightly superior results.
Luc W.F. van Haaster, Erem Tutal, R. Mekary et al.· Brain and Spine· 0 citations
Background/Objectives: This study aimed to evaluate whether artificial intelligence-derived vertebral volumetric bone mineral density (AI-vBMD) and paraspinal intermuscular adipose tissue (IMAT) ratio from routine computed tomography (CT) could identify moderate-to-severe vertebral compression fractures (VCFs) in breast cancer survivors, and whether paraspinal IMAT ratio and routinely available clinical variables improved diagnostic performance. Methods: This retrospective study included 275 women with breast cancer who underwent routine non-contrast CT and lumbar quantitative computed tomography (QCT). Hounsfield unit-derived volumetric bone mineral density (HU-vBMD) was derived using a QCT-referenced HU-to-vBMD conversion equation, whereas AI-vBMD and paraspinal IMAT ratio were extracted using automated software. Moderate-to-severe VCF was defined as Genant grade ≥ 2. Agreement with QCT-vBMD was assessed using correlation, intraclass correlation coefficient (ICC), and Bland–Altman analysis. Model discrimination was evaluated using receiver operating characteristic analysis and DeLong tests. Results: Moderate-to-severe VCF was present in 75 patients (27.3%). HU-vBMD and AI-vBMD showed excellent agreement with QCT-vBMD (ICC, 0.978 and 0.987, respectively). AI-vBMD outperformed HU-vBMD for identifying VCFs (AUC, 0.738 vs. 0.714; p < 0.001). IMAT ratio showed comparable standalone discrimination to AI-vBMD (AUC, 0.760 vs. 0.738; p = 0.604). Adding IMAT ratio to AI-vBMD improved discrimination (AUC, 0.786 vs. 0.738; p = 0.038). The full model incorporating clinical covariates achieved the highest AUC (0.828; 95% CI, 0.778–0.878). Conclusions: AI-vBMD and paraspinal IMAT ratio automatically extracted from routine CT improved the diagnostic assessment of prevalent moderate-to-severe VCFs in breast cancer survivors. This study supports an automated CT-based approach that integrates vertebral bone density and paraspinal muscle–fat information for opportunistic identification of clinically relevant VCFs.
Chen Wan, Lingquan Kong, Jie Hao et al.· Journal of Clinical Medicine· 0 citations
Study Design
Retrospective study.
Purpose
To evaluate the influence of lipid parameters on vertebral bone quality (VBQ) scores in patients undergoing lumbar fusion surgery.
Overview of Literature
The MRI-derived VBQ score is a validated tool for assessing osteoporosis in patients undergoing lumbar fusion surgery. However, the effect of lipid parameters on its diagnostic performance remains unclear.
Methods
Consecutive patients who underwent lumbar fusion surgery over a 3-year period were retrospectively reviewed. Inclusion required the availability of lumbar T-scores, quantitative computed tomography, magnetic resonance imaging (MRI), and laboratory data obtained within a 1-month interval. Patients were stratified according to lipid levels and clinical characteristics. Correlation analysis and receiver operating characteristic (ROC) curve analysis evaluated the impact of lipid parameters on VBQ 1.5T, VBQ 3.0T, and the calibrated VBQ (C-VBQ) scores, which was developed to improve consistency across different MRI field strengths.
Results
This study included 294 patients (mean age, 67.1±6.5 years; 57.8% female; 56.8% with hyperlipidemia). Among the lipid parameters evaluated, only high-density lipoprotein cholesterol (HDLC) demonstrated a modest association with VBQ scores. Correlation analysis demonstrated significant positive correlations between HDLC and both VBQ 1.5T (p =0.003) and C-VBQ (p =0.001). After stratification by HDLC levels, C-VBQ differed significantly between normal and abnormal HDLC groups (p =0.046). Significant differences in VBQ 3.0T and C-VBQ across HDLC quintiles persisted even after T-score adjustment (p =0.037 and p =0.003). ROC analysis revealed minimal influence of lipid parameters or hyperlipidemia on VBQ diagnostic performance.
Conclusions
Among lipid parameters, only HDLC was positively associated with VBQ scores; however, this weak association did not significantly compromise the diagnostic performance of VBQ for osteoporosis.
Fangke Hu, Dapeng Yu, Penghe Li et al.· Asian Spine Journal· 0 citations
BACKGROUND/OBJECTIVES
Osteoporosis is a systemic skeletal disease characterized by low bone mineral density (BMD) and increased fracture risk. Dual-energy X-ray absorptiometry (DXA) is the reference standard for diagnosis, typically assessing the lumbar spine and femoral neck. Site-specific differences may result in spine-hip BMD discordance, potentially influencing clinical decision making. This study aimed to determine the prevalence, patterns, and predictors of spine-hip BMD discordance among Jordanian adults.
METHODS
This retrospective observational study included adults aged ≥18 years who underwent DXA for osteoporosis evaluation at two tertiary hospitals in Jordan between March 2024 and June 2025. Lumbar spine and femoral neck BMD were measured using standardized DXA protocols. Discordance was classified using T-scores or Z-scores as appropriate.
RESULTS
A total of 826 participants were included, comprising 576 evaluated using T-scores and 250 using Z-scores. In the T-score group, lumbar spine-femoral neck BMD discordance was observed in 66.0% of participants, including 50.5% with minor discordance and 15.5% with major discordance, while 34.0% were concordant. In contrast, discordance was present in only 18.4% of the Z-score group. After adjusting for principal demographic and high-prevalence clinical parameters, Multivariable analysis identified family history of osteoporosis and sex as independent predictors of discordance severity in the T-score group, whereas no significant predictors were identified in the Z-score group.
CONCLUSIONS
Spine-hip BMD discordance is common among Jordanian adults undergoing DXA, particularly in those assessed using T-scores, underscoring the importance of multisite DXA assessment to improve diagnostic accuracy and fracture risk evaluation.
Mais Alqoudah, Mu’taz Massad, Zain aldeen Saleh et al.· Journal of clinical densitom...· 0 citations