The Association between opportunistic DECT-derived bone values and vertebral fracture status

Le résumé fourni par la source

INTRODUCTION: Osteoporosis is a major cause of spinal insufficiency fractures but often remains underdiagnosed. Dual-energy CT (DECT) enables reliable assessment of bone mineral density (BMD), but its limited accessibility highlights the need for alternative metrics. This study investigates the association between vertebral fracture status, DECT-derived BMD and alternative bone quality assessments, including Hounsfield Unit (HU)-based assessment and cortical thickness ratio. METHODS: A total of 180 patients who underwent non-contrast DECT of the spine between January 2016 and December 2021 were retrospectively analyzed. All imaged vertebrae were assessed for acute insufficiency fractures. DECT-based BMD was assessed using a dedicated postprocessing software that applies material decomposition and compared with HU-based assessment and cortical thickness ratio. Statistical analysis included correlation analysis, logistic regression adjusted for age and sex, ROC and PR curve analyses. RESULTS: , p < 0.001), trabecular HU (93.5 vs. 159.5 HU, p < 0.001), and cortical thickness ratio (1.065 vs. 1.05, p < 0.001). Cortical HU showed no significant difference between patients with and without fractures (p = 0.35). DECT-based BMD provided the highest diagnostic accuracy for detecting insufficiency fractures, yielding an AUC of 0.8 for the ROC curve and an AUC of 0.9 for the PR curve. HU-based measurements (trabecular HU: Spearman ρ = 0.17; cortical HU: ρ = 0.2) and the cortical thickness ratio (ρ = -0.01) demonstrated only weak correlations with the reference standard, DECT-derived BMD. CONCLUSION: DECT-based BMD demonstrated the highest diagnostic accuracy for insufficiency fractures. HU-based assessments and cortical thickness ratio showed only weak correlations with DECT-based BMD, limiting their reliability as alternatives.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.