Identification of Bone Marrow Edema and Osteochondral Injuries of the Knee: Diagnostic Accuracy of Dual-energy CT and Virtual Non-calcium Techniques

Purpose: To evaluate prospectively the diagnostic accuracy values of dual-energy computed tomography (DECT) to identify bone marrow edema and osteochondral injuries of the knee.

Methods and Materials: This prospective study approved by the institutional review board included 33 consecutive patients (20 men and 13 women; mean age: 54.2 years [range: 41–76 years]) studied between March 2017 and March 2018. All patients underwent DECT (80 kV and tin filter 150 kV) and magnetic resonance imaging (MRI) with short tau inversion recovery (STIR) images within 7 days. DECT data were postprocessed on a dedicated offline workstation (syngo.via VB20; Siemens, Erlangen, Germany) by using a three-material decomposition algorithm for generating non-calcium images of the knee. Two radiologists (25 and 11 years of experience, respectively), blinded to clinical data, evaluated the presence of abnormal attenuation of each knee on dedicated color-coded maps and on gray-scale images. STIR images served as the standard of reference. Diagnostic accuracy values of the DECT maps and of the CT numbers (quantitative assessment) by using receiver operator curves (ROC) and relative area under the curve (AUC) were calculated. Interobserver and intraobserver agreements were calculated with κ statistics. Continuous and categorical variables were evaluated by using the t test and chi-square or Fisher exact test, as appropriate. A value of p < 0.05 was considered statistically significant.

Results: MRI revealed the presence of bone marrow edema of the knee in 17 of 33 cases (51.6%), with 10 osteochondral injuries. DECT numbers were significantly different between positive (mean − 5.5 ± 32.6 HU) and negative cases (mean − 72.4 ± 41.6 HU) with a p value < 0.001. The ROC curve analysis revealed an AUC of 0.904 (95% confidence interval, 0.776–0.948). By using a − 15 HU cutoff to identify bone marrow edema, the sensitivity, specificity, positive predictive value, and negative predictive value, and accuracy of DECT were 89.5%, 92.9%, 94.4%, 86.6%, and 90.9%, respectively. The interobserver and intraobserver agreement were near perfect (κ = 0.82 and κ = 0.86, respectively).

Conclusion: DECT represents a reliable imaging tool for demonstration of bone marrow edema and osteochondral injuries of the knee.