|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Vascular and Interventional Radiology |
1 From the Institute of Diagnostic Radiology (P.S., T.F., T.P., N.P., H.S., B.M., H.A., T.S.) and Clinic for Cardiovascular Surgery (M.L.), University Hospital Zurich, Raemistrasse 100, 8091 Zurich, Switzerland; and Siemens Medical Solutions, Forchheim, Germany (B.S.). Received January 30, 2008; revision requested March 28; revision received April 5; accepted April 8; final version accepted April 10. Supported by the National Center of Competence in Research, Computer Aided and Image Guided Medical Interventions of the Swiss National Science Foundation. Address correspondence to H.A. (e-mail: hatem.alkadhi{at}usz.ch).
Purpose: To assess the diagnostic performance of dual-energy dual-source computed tomography (CT) in the detection of endoleaks after endovascular abdominal aortic aneurysm (AAA) repair.
Materials and Methods: This study was local ethics board approved, and written informed consent was obtained from all patients. One hundred eighteen patients (21 women, 97 men; mean age, 74 years ± 8 [standard deviation]) underwent follow-up dual-energy dual-source CT during the nonenhanced, arterial, and delayed phases after AAA repair. Delayed phase CT images were acquired in the dual-energy mode for reconstruction of virtual nonenhanced images. Two blinded and independent readers evaluated the data for the presence or absence of endoleaks during three reading sessions: Standard nonenhanced, arterial phase, and delayed phase images were read during session A; virtual nonenhanced, arterial phase, and delayed phase images, during session B; and virtual nonenhanced and delayed phase images, during session C. Sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) were calculated, with the session A image data set as the reference standard. Radiation dose parameters were estimated.
Results: Reading session A revealed that 52 (44%) of 118 patients had endoleaks. Overall sensitivity, specificity, NPV, and PPV for CT endoleak detection during sessions B and C were identical: 100%, 97%, 100%, and 96%, respectively. The accuracy of the session B and session C readings was not significantly different from that of the session A reading (P = .50). The effective radiation dose in the image acquisition protocol involving one dual-energy scan was significantly (P < .001) lower than the effective doses in the protocols involving standard triple-phase scanning (mean difference, 61%) and standard nonenhanced and delayed phase scanning (mean difference, 41%).
Conclusion: Compared with standard protocols, one dual-energy dual-source CT scan performed during the delayed phase with reconstruction of virtual nonenhanced images enables detection of endoleaks after endovascular AAA repair with high accuracy and a considerably lower radiation dose.
© RSNA, 2008
Related Article
Radiology 2008 249: 692-700.
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |