Skip to Content
DiagnosticsDiagnostics
  • Article
  • Open Access

19 November 2026

16 Pages

Development of a One-Stop Bone Mineral Density Estimation Model Based on Dual-Energy CT with Quantitative Computed Tomography as the Reference Standard

,
,
and
1
Department of Radiology, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou 310000, China
2
Department of Radiology, the First Affiliated Hospital of Dalian Medical University, Dalian 116000, China
*
Author to whom correspondence should be addressed.
Diagnostics2026, 16(22), 3890;https://doi.org/10.3390/diagnostics16223890 
(registering DOI)
This article belongs to the Section Medical Imaging and Theranostics

Abstract

Background: This study intended to explore potential diagnostic criteria for bone status evaluation using a dual-energy computed tomography (DECT) material decomposition technique through comparison with quantitative computed tomography (QCT). Methods: In total, 355 patients who underwent abdominal unenhanced and contrast-enhanced (arterial-, venous- and delayed-phase) DECT were randomly divided into training group A with 255 cases and validation group B with 100 cases (sex was coded as 1 for male and 2 for female). Group A was further divided into four groups, a, b, c, d, according to age (<30 years, 30–49 years, 50–69 years, ≥70 years). QCT standard scanning and reconstruction parameters were adopted for the unenhanced phase, while spectral imaging mode was applied for all enhanced phases. The hydroxyapatite (iodine) density (DHAP (iodine)) in the trabecular bone of L1–L3 vertebral bodies was measured and averaged, alongside QCT-derived bone mineral density (BMD). Spearman’s correlation test was used to analyze the relationship between DECT and QCT BMD measurements. DHAP (iodine) of osteoporosis, osteopenia and normal vertebral bodies were compared by one-way ANOVA. ROC curves were constructed to determine the optimal diagnostic thresholds of DHAP (iodine) for differentiating the three bone status categories. The correlated fitting curves between DHAP (iodine) and two variables (age and sex) were established separately for each subgroup in group A. The regression formula from group A was used to estimate BMD in group B. The differences between measured and formula-calculated BMD in group B were assessed. Results: DHAP (iodine) of L1–L3 vertebral bodies showed no significant differences across the three contrast-enhanced CT phases (all p > 0.05). In group A, DHAP (iodine) was positively related to QCT BMD (r = 0.968, p < 0.001). The differences in DHAP (iodine) among different age groups were statistically significant (p < 0.001) with no correlation to age in group a (p > 0.05), but negative correlations with age in groups b, c and d (r = −0.399, −0.489, −0.397, all p < 0.05). No significant sex-related correlation was identified in groups a, b and d (all p > 0.05), while a negative correlation with sex was present in group c (r = −0.233, p < 0.05). For patients aged 30 years and above, DHAP (iodine) showed a strong overall inverse correlation with age (r = −0.731, p < 0.001), with no significant correlation between DHAP (iodine) and sex (p > 0.05): DHAP (iodine) = −1.053 * age + 755.517 (2 mg/cm3). There were significant differences in DHAP (iodine) among osteoporosis, osteopenia and normal groups (p < 0.001). For the diagnosis of osteopenia using the threshold DHAP (iodine) ≤694.89 (2 mg/cm3), the area under the ROC curve, sensitivity and specificity were calculated as 0.990, 93.50% and 96.60%. The corresponding values for identifying osteoporosis were 0.962, 90.30%, 90.90% and with DHAP (iodine) ≤677.93 (2 mg/cm3), respectively. In group B, formula-calculated BMD showed high agreement with measured BMD. Conclusions: The one-stop DECT-based estimation model for vertebral BMD quantification shows technical feasibility and promising diagnostic accuracy.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.