Identification of bone marrow edema of the ankle: diagnostic accuracy of dual-energy CT in comparison with MRI.

Abstract

To evaluate the diagnostic accuracy of dual-energy computed tomography (DECT) to identify bone marrow edema (BME) of the ankle. This prospective institutional review board approved study included 40 consecutive patients (29 males and 11 females; mean age of 32.3years) studied with DECT (80kV and tin filter 150kV) and MRI within 10days. DECT data were post-processed for generating non-calcium images of the ankle. Two radiologists evaluated the presence of BME on color-coded maps. Diagnostic accuracy values for diagnosing BME were calculated for DECT maps (qualitative assessment) and for CT numbers (quantitative assessment) by using receiver operator curves and area under the curve (AUC), using MRI as the gold standard. Interobserver and intraobserver agreements were calculated with k-statistics. A p < 0.05 was considered statistically significant. DECT depicted BME in 23/25 patients (92.0%). The sensitivity, specificity, PPV and NPV, and accuracy achieved by evaluating the DECT images were 92.0, 86.6, 92.0, 84.6, and 90.0%, for reader 1 and 88.0, 86.6, 91.6, 78.6, and 87.5, for reader 2, respectively. The interobserver and intraobsever agreements were near perfect (k = 0.87 and k = 0.83, respectively). DECT numbers were significantly different between positive (mean - 12.6 ± 29.6 HU) and negative cases (mean - 64.2 ± 34.5 HU) with a p value < 0.001. By using - 20HU cutoff to identify BME (AUC of 0.896.), the sensitivity, specificity, PPV and NPV, and accuracy of the quantitative analysis were 88.0, 92.6, 95.7, 92.6, and 87.5%, respectively. DECT represents an accurate imaging tool for demonstration of BME of the ankle when compared to MRI.