MRI-Based Vertebral Bone Quality Score for QCT-Defined Osteoporosis: Incremental Diagnostic Value and an Exploratory 1.5-T Threshold in a Retrospective Study

Background: Osteoporosis is frequently underrecognized in patients undergoing lumbar magnetic resonance imaging (MRI), creating an opportunity for imaging-based case finding. Methods: This retrospective diagnostic accuracy study evaluated the MRI-based vertebral bone quality (VBQ) score against quantitative computed tomography (QCT)-defined osteoporosis, its incremental diagnostic value beyond age and sex, and the stability of a threshold derived under a standardized 1.5-T protocol. Among 281 consecutive adults, 278 formed the primary MRI cohort and 234 with valid L1–L2 QCT measurements formed the reference-standard cohort; 51 had osteoporosis. VBQ was calculated from L1–L4 T1-weighted vertebral signal intensities normalized to L3 cerebrospinal fluid intensity. Results: Interobserver reliability was high (intraclass correlation coefficient (ICC(2,2)), 0.919; 95% confidence interval (CI), 0.899–0.935). VBQ correlated inversely with QCT volumetric bone mineral density (r = −0.687; p < 0.001) and remained independently associated with osteoporosis after adjustment for age and sex (adjusted OR, 6.08; 95% CI, 2.36–15.68). Adding VBQ increased the apparent model area under the receiver operating characteristic curve (AUC) from 0.847 to 0.896 (bootstrap optimism-corrected AUCs, 0.839 and 0.887). Standalone AUCs were 0.833 for VBQ and 0.863 for fat fraction. A VBQ threshold >3.7047 yielded 74.51% sensitivity and 83.06% specificity, but bootstrap-selected thresholds varied (2.5th–97.5th percentiles, 3.409–3.851). Conclusion: VBQ may provide an adjunctive opportunistic marker on routine lumbar MRI. However, its threshold is exploratory and requires external validation; VBQ should not be used alone to exclude osteoporosis.