Diagnostic accuracy of dual-energy CT and virtual non-calcium techniques to evaluate bone marrow edema in vertebral compression fractures.

The aim of this study was to evaluate DECT diagnostic accuracy in the identification of vertebral bone marrow edema, using MRI as standard of reference. This prospective institutional review board-approved study included 76 consecutive patients (29 males and 47 females; mean age 62.3, range 51-82 years) studied with DECT (90 kV and tin filter 150 kV) and MRI within 7 days. Three radiologists evaluated DECT (reader 1 and 2) and MRI images (reader 3). Diagnostic accuracy of the DECT maps (qualitative assessment) and of the CT numbers (quantitative assessment), interobserver and intraobserver agreements were calculated. MRI revealed 61 edematous vertebrae and 52 collapsed non-edematous vertebrae. The sensitivity, specificity, PPV and NPV and accuracy of the qualitative assessment of the DECT maps were 88.6, 92.3, 93.1, 87.3 and 90.3%, for reader 1, 90.2, 90.3, 91.6, 88.7 and 90.3, for reader 2, and 91.8, 90.4, 91.6, 90.4 and 91.1% for quantitative analysis, respectively. DECT numbers were significantly different between positive (mean - 23 HU, range - 189, 29 HU) and negative cases (mean - 126 HU, range - 321, - 66 HU) with p < 0.001. The ROC curve analysis revealed an AUC of 0.886 (95% confidence interval 0.722-0.913). The interobserver and intraobserver agreements were near perfect (k = 0.87 and k = 0.83, respectively). DECT represents an accurate imaging technique for demonstrating bone marrow edema in vertebral compression fracture, if compared to MRI.