Acad Radiol. 2022 Nov 9:S1076-6332(22)00572-4. doi: 10.1016/j.acra.2022.10.016. Online ahead of print.
ABSTRACT
RATIONALE AND OBJECTIVES: We investigated the diagnostic performance of dual-energy CT (DECT) virtual non-calcium (VNCa) and Rho/Z images for bone marrow infiltration of primary malignant bone tumors (PMBTs).
MATERIALS AND METHODS: We retrospectively analyzed 65 patients with PMBT who underwent DECT and MRI within 2 weeks. DECT was used to evaluate the presence and extent of marrow involvement surrounding PMBTs using the SCT, VNCa, and Rho/Z images. MRI was used as the reference standard for measurements. CT values of normal and involved bone marrow areas were measured on VNCa images, and Zeff values were measured on Rho/Z images. The statistical methods used were the 2*C chi-square test, ANOVA test, paired samples t test, and diagnostic performance of the different variables were evaluated using receiver operating characteristic curves.
RESULTS: VNCa and Rho/Z images showed higher accuracy (91%, 92% vs. 67%) and sensitivity (90%, 92% vs. 69%) than SCT images for diagnosing bone marrow infiltration in patients with PMBT. The maximum longitudinal diameter of tumor involvement measurements was statistically different between VNCa and SCT, Rho/Z and SCT, MRI, and SCT (all p < 0.05, p = 0.047, p = 0.049, and p = 0.023, respectively). The maximum transverse diameter was statistically significant between SCT and MRI, VNCa and MRI, Rho/Z and MRI (all p < 0.05, and p = 0.015, and p = 0.044, and p = 0.047, respectively). The HU or Zeff values based on the area of interest of VNCa and Rho/Z images differed significantly between the normal and infiltrated bone marrow area (p < 0.001). Receiver operating characteristic curve analysis revealed area under the curves of 0.995 and 0.988, respectively, with cut-off values of -31.57 HU and 7.8, and the sensitivity of both was 96.9%.
CONCLUSION: DECT-VNCa and Rho/Z images have good diagnostic value when evaluating bone marrow infiltration in PMBTs.
PMID:36371375 | DOI:10.1016/j.acra.2022.10.016
