COMPUTED TOMOGRAPHY Impact of dual-energy CT post-processing to differentiate venous thrombosis from iodine flux artefacts Jakob Weiss 1 & Christoph Schabel 1 & Ahmed E. Othman 1 & Georg Bier 2 & Konstantin Nikolaou 1 & Fabian Bamberg 1 & Malte N. Bongers 1 Received: 27 February 2018 /Revised: 26 April 2018 /Accepted: 11 May 2018 # European Society of Radiology 2018 Abstract Objectives To investigate the accuracy of dual-energy (DE) CT-based iodine maps (IM) and noise-optimised monoenergetic extrapolations (MEI+) at 40 keV for the detection and differentiation of venous thrombosis (VT) from iodine flux artefacts (IFA) in comparison to portal-venous phase CT (CT PV ). Methods Ninety-nine patients were enrolled in this study. In all patients, VT or IFA was suspected on contrast-enhanced CT and confirmed by follow-up CT or colour-coded ultrasound. All examinations were performed on a third-generation dual-source CT system in DE mode during portal-venous phase. CT PV , IM and 40-keV MEI+ were reconstructed and independently evaluated by two radiologists for the presence/absence of VT and/or IFA. Diagnostic confidence was rated on a three-point scale (3 = high confidence). Quantitative parameters were obtained by calculating contrast-to-noise ratios (CNRs), iodine content and thrombus volume. Diagnostic accuracy was assessed by calculating receiver operating characteristics (ROC) of CNR. Results Diagnostic confidence was significantly higher for IM and MEI+ [both 3 (2-3)] compared to CT PV [2 (1-3); p 0.03]. ROC analysis revealed significantly higher AUC values and increased sensitivity for IM and MEI+ (AUC = 88%/sensitivity = 79.1% and 86%/73.1%) than for CT PV (75%/61.2%; p 0.01). Thrombus volume was significantly higher in MEI+ than in IM and CT PV (p < 0.001). CNR of thrombosis was significantly higher in IM [11.5 (8.5-14.5), p < 0.001) and MEI+ [10.9 (8.8-15.5), p < 0.001] than in CT PV [8.2 (5.8-11.4)]. Iodine quantification revealed significantly lower results in VT than in IFA [0.55 mg/ml (0.23-0.90) and 1.81 (1.60-2.12) mg/ml; p < 0.001]. Conclusions IM and MEI+ 40 keV showed significantly higher diagnostic confidence and accuracy for the detection and differentiation of VT from IFA in comparison to CT PV . Key Points Iodine maps and noise-optimised monoenergetic extrapolations at 40 keV increase diagnostic confidence and accuracy for the detection and differentiation of venous thrombosis from iodine flux artefacts. Dual-energy post-processing can significantly increase contrast-to-noise ratio and the sensitivity for the diagnosis of venous thrombosis Iodine load in venous thrombosis is significantly lower than in iodine flux artefacts Keywords Venous thrombosis . Diagnostic imaging . Contrast media Abbreviations ADMIRE Advanced modelled iterative reconstruction algorithm CT PV Portal-venous phase computed tomography IFA Iodine flux artefact IM Iodine maps MEI+ Monoenergetic extrapolations VT Vein thrombosis * Christoph Schabel Christoph.Schabel@med.uni-tuebingen.de 1 Department of Diagnostic and Interventional Radiology, University Hospital of Tübingen, Hoppe-Seyler-Strasse 3, 72076 Tübingen, Germany 2 Department of Diagnostic and Interventional Neuroradiology, University Hospital of Tübingen, Hoppe-Seyler-Strasse 3, 72076 Tübingen, Germany European Radiology https://doi.org/10.1007/s00330-018-5534-8