|| Checking for direct PDF access through Ovid
This study examined the ability of QCT-based structural assessment techniques to predict vertebral strength in anterior bending.The purpose of this study was to compare the abilities of QCT-based bone mineral density (BMD), mechanics of solids models (MOS), e.g., bending rigidity, and finite element analyses (FE) to predict the strength of isolated vertebral bodies under anterior bending boundary conditions.Although the relative performance of QCT-based structural measures is well established for uniform compression, the ability of these techniques to predict vertebral strength under nonuniform loading conditions has not yet been established.Thirty human thoracic vertebrae from 30 donors (T9–T10, 20 female, 10 male; 87 ± 5 years of age) were QCT scanned and destructively tested in anterior bending using an industrial robot arm. The QCT scans were processed to generate specimen-specific FE models as well as trabecular bone mineral density (tBMD), integral bone mineral density (iBMD), and MOS measures, such as axial and bending rigidities.Vertebral strength in anterior bending was poorly to moderately predicted by QCT-based BMD and MOS measures (R2 = 0.14–0.22). QCT-based FE models were better strength predictors (R2 = 0.34–0.40); however, their predictive performance was not statistically different from MOS bending rigidity (P > 0.05).Our results suggest that the poor clinical performance of noninvasive structural measures may be due to their inability to predict vertebral strength under bending loads. While their performance was not statistically better than MOS bending rigidities, QCT-based FE models were moderate predictors of both compressive and bending loads at failure, suggesting that this technique has the potential for strength prediction under nonuniform loads. The current FE modeling strategy is insufficient, however, and significant modifications must be made to better mimic whole bone elastic and inelastic material behavior.