Pulmonary embolism detection with dual-energy CT: experimental study of dual-source CT in rabbits.

PURPOSE: To evaluate feasibility and added value of dual-energy computed tomography (CT) in diagnosis of pulmonary embolism (PE).

MATERIALS AND METHODS: This institutional animal experimental committee-approved study was performed in accordance with animal care guidelines. Eight New Zealand rabbits underwent standard unenhanced and contrast material-enhanced dual-source CT. Gelatin sponge particles were injected into the pulmonary artery, and rabbits underwent contrast-enhanced dual-source CT pulmonary angiography, from which blood-flow (BF) and fusion images were created. Immediately after dual-source CT, rabbits were sacrificed, their lungs were removed and fixed in 10% formalin, and detailed pathologic determination of location and number of lung lobes with PE was performed. Two rabbits were excluded: One died during the procedure. In the other, the catheter tip was retained in the left inferior pulmonary artery. This caused marked postembolization CT image artifacts in adjacent regions. Six rabbits were included in final analysis. Two radiologists without knowledge of pathologic results evaluated five pulmonary lobes in each rabbit and recorded whether PE was present. Pathologic results served as the reference standard. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the techniques were calculated. Weighted kappa values were calculated to evaluate agreement between modalities.

RESULTS: Pathologic analysis revealed PE in 18 of 30 pulmonary lobes. Conventional CT angiography was used to correctly identify PE in 12 lobes and absence of emboli in 18 lobes, which corresponded to sensitivity, specificity, PPV, and NPV of 67%, 100%, 100%, and 67%, respectively. A kappa value of 0.65 indicated good correlation with pathologic findings. On BF images, segments with an embolic region showed low perfusion compared with segments with a normal pulmonary region. BF images and fused images correctly showed PE in 16 of 18 pulmonary lobes and absence of emboli in 11 of 12 lobes, which corresponded to sensitivity, specificity, PPV, and NPV of 89%, 92%, 94%, and 85%, respectively, in detection of PE. A kappa value of 0.80 indicated good correlation with pathologic findings.

CONCLUSION: Dual-source CT can depict normal and abnormal blood perfusion distribution in a rabbit's lung. Abnormal pulmonary blood distribution, as shown at dual-source CT, improves detection of acute PE in rabbits.