Dual-energy CT angiography-derived virtual non-contrast images for follow-up of patients with surgically clipped aneurysms: a retrospective study.

Abstract

To evaluate the diagnostic performance, image quality, artifacts, and radiation doses of virtual non-contrast (VNC) images, relative to true non-contrast (TNC) images, in patients with surgically clipped aneurysms. Seventy-six patients who underwent unenhanced computed tomography (CT) and dual-energy (DE)-CT angiography after surgical clipping of 85 intracranial aneurysms were included in the study. Diagnostic performances of VNC and TNC images were compared with respect to sensitivity, specificity, and positive and negative predictive values. The results of quantitative and qualitative analyses were compared between VNC and TNC images. Radiation doses were also compared between VNC and TNC images. Diagnostic performance of VNC images was lower than that of TNC images; however, addition of contrast-enhanced images improved diagnostic performance. Image noise and mean attenuation of VNC images were significantly higher than those of TNC images in the centrum semiovale, cerebellum, and pons. The quality of VNC images was excellent or sufficient (85.5% for observer 1, 85.5% for observer 2), and complete acceptance of VNC images was achieved in 64.5% for observer 1 and in 71.0% for observer 2; however, the addition of contrast-enhanced images increased the level of acceptance (92.0% for observer 1, 90.9% for observer 2). Clip artifacts were significantly lower in VNC images than in TNC images. CT dose index, dose-length product, and effective dose were significantly lower without TNC images. VNC images showed lower diagnostic performance and image quality, and higher image noise than TNC images; however, VNC images could reduce clip artifacts and radiation doses.