Objective <p>To compare the qualitative diagnostic performance of conventional computed tomography (CT) alone with that of dual-energy computed tomography (DECT) virtual non-calcium (VNCa) imaging for detecting spinal metastases, and evaluate the diagnostic utility of DECT-derived quantitative material parameters in this retrospective diagnostic accuracy study.</p> Materials and methods <p>Consecutive patients (August 2019–January 2024) underwent DECT at a single center. Two board-certified radiologists scored vertebrae on conventional CT and CT + VNCa using a 5-point scale (positive for scores 4–5). The reference standard was a consensus MRI within 3&#xa0;months, supplemented by PET/CT and/or bone scintigraphy when available. We estimated sensitivity, specificity, and AUC with 95% CIs. Quantitative parameters including 70-keV CT attenuation (HU), water–calcium, water–hydroxyapatite, and <i>Z</i><sub>eff</sub> were compared between metastatic and normal vertebrae using the Wilcoxon rank-sum test.</p> Results <p>We included 117 patients, comprising 1416 vertebrae; 24.2% had metastases. For conventional CT, sensitivity was 0.70/0.71 (95% CI 0.664–0.738/0.677–0.750) and specificity, 0.97/0.97 (95% CI 0.939–1.000/0.935–1.000). CT + VNCa sensitivity increased to 0.94/0.93 (95% CI 0.906–0.980/0.890–0.964; <i>p</i> &lt; 0.001) and AUC improved to 0.97/0.96 (both <i>p</i> &lt; 0.001, specificity, 0.99/0.98). Quantitative analysis showed significantly higher values of water–calcium (1133.8 vs. 1007.6&#xa0;mg/cm<sup>3</sup>), water–hydroxyapatite (991.5 vs. 971.0&#xa0;mg/cm<sup>3</sup>), and effective atomic number (10.5 vs. 8.3) in osteoblastic metastases compared with those in normal vertebrae (all <i>p</i> &lt; 0.05).</p> Conclusion <p>CT + VNCa significantly improves the detection of spinal metastases compared with conventional CT alone. Quantitative parameters from material decomposition analysis provide additional diagnostic value.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Diagnostic accuracy of dual-energy CT virtual non-calcium imaging for detecting spinal metastases: a retrospective comparison with conventional CT

  • Hirotaka Nakashima,
  • Junji Mochizuki,
  • Syunsuke Itaya,
  • Hiroyuki Yoshino,
  • Yoshinao Kodama,
  • Yasuo Sakurai,
  • Takanori Masuda,
  • Kazuhiro Sato,
  • Yuji Yaegashi

摘要

Objective

To compare the qualitative diagnostic performance of conventional computed tomography (CT) alone with that of dual-energy computed tomography (DECT) virtual non-calcium (VNCa) imaging for detecting spinal metastases, and evaluate the diagnostic utility of DECT-derived quantitative material parameters in this retrospective diagnostic accuracy study.

Materials and methods

Consecutive patients (August 2019–January 2024) underwent DECT at a single center. Two board-certified radiologists scored vertebrae on conventional CT and CT + VNCa using a 5-point scale (positive for scores 4–5). The reference standard was a consensus MRI within 3 months, supplemented by PET/CT and/or bone scintigraphy when available. We estimated sensitivity, specificity, and AUC with 95% CIs. Quantitative parameters including 70-keV CT attenuation (HU), water–calcium, water–hydroxyapatite, and Zeff were compared between metastatic and normal vertebrae using the Wilcoxon rank-sum test.

Results

We included 117 patients, comprising 1416 vertebrae; 24.2% had metastases. For conventional CT, sensitivity was 0.70/0.71 (95% CI 0.664–0.738/0.677–0.750) and specificity, 0.97/0.97 (95% CI 0.939–1.000/0.935–1.000). CT + VNCa sensitivity increased to 0.94/0.93 (95% CI 0.906–0.980/0.890–0.964; p < 0.001) and AUC improved to 0.97/0.96 (both p < 0.001, specificity, 0.99/0.98). Quantitative analysis showed significantly higher values of water–calcium (1133.8 vs. 1007.6 mg/cm3), water–hydroxyapatite (991.5 vs. 971.0 mg/cm3), and effective atomic number (10.5 vs. 8.3) in osteoblastic metastases compared with those in normal vertebrae (all p < 0.05).

Conclusion

CT + VNCa significantly improves the detection of spinal metastases compared with conventional CT alone. Quantitative parameters from material decomposition analysis provide additional diagnostic value.