Simplified sagittal slice HU measurement as a practical alternative to conventional axial HU in opportunistic osteoporosis screening
Akihiko Hiyama, Daisuke Sakai, Hiroyuki Katoh, Masato Sato, Masahiko Watanabe
Abstract
PURPOSE: Opportunistic assessment of bone quality using CT-derived Hounsfield Unit (HU) values has gained attention as an alternative to DXA-based osteoporosis screening. However, conventional axial HU measurements are time-consuming and operator-dependent. A simplified sagittal HU measurement method may provide a more practical approach, although its clinical validity has yet to be fully established. This study aimed to evaluate the correlation between HU values obtained from simplified sagittal and conventional axial slices and determine their respective diagnostic performance in identifying osteoporosis. METHODS: We retrospectively analyzed 162 patients aged ≥ 50 years who underwent both lumbar spine CT and DXA scans. HU values were measured using both conventional axial slices (Ax L1-4 Ave HU) and a simplified sagittal slice method (Sg L1-4 Ave HU). Spearman's rank correlation was calculated to assess the association between each HU measurement and the lowest T-score (Low-T score). Agreement between the two methods was evaluated using Bland-Altman analysis, and the predictive performance for diagnosing osteoporosis (Low T-score ≤ - 2.5) was assessed by ROC curve analysis. Vertebra-specific differences between axial and sagittal HU values were also analyzed. RESULTS: Sg L1-4 Ave HU showed a moderate correlation with Low T-score (ρ = 0.476), comparable to Ax L1-4 Ave HU (ρ = 0.493). The Bland-Altman analysis revealed a mean difference of - 4.08 HU between Sg and Ax measurements, with most differences falling within the 95% limits of agreement. Axial HU values were consistently higher than sagittal HU values across all lumbar levels, particularly at L1 and L4. The AUC for predicting osteoporosis was 0.692 for Sg L1-4 Ave HU and 0.707 for Ax L1-4 Ave HU. Based on Youden's index, the optimal cutoff values were 94.0 HU for sagittal and 99.5 HU for axial measurements. CONCLUSIONS: Simplified sagittal HU measurements demonstrated good agreement with conventional axial measurements and exhibited similar diagnostic accuracy in identifying osteoporosis. Given their efficiency and consistency, sagittal HU values may serve as a practical and reproducible alternative for CT-based opportunistic bone quality assessment, especially in routine clinical settings or where DXA is unavailable.