Abdominal vessel depiction on virtual triphasic spectral detector CT: initial clinical experience

PURPOSE: To evaluate vessel assessment in virtual monoenergetic images (VMI(40keV)) and virtual-non-contrast images (VNC) derived from venous phase spectral detector computed tomography (SDCT) acquisitions in comparison to arterial phase and true non-contrast (TNC) images. METHODS: Triphasic abdominal SDCT was performed in 25 patients including TNC, arterial and venous phase. VMI(40keV) and VNC were reconstructed from the venous phase and compared to conventional arterial-phase images (CI(art)), TNC and conventional venous-phase images (CI(ven)). Vessel contrast and virtual contrast removal were analyzed with region-of-interest-based measurements and in a qualitative assessment. RESULTS: Quantitative analysis revealed no significant attenuation differences between TNC and VNC in arterial vessels (p-range 0.07–0.47) except for the renal artery (p = 0.011). For venous vessels, significant differences between TNC and VNC were found for all veins (p < 0.001) except the inferior vena cava (p = 0.26), yet these differences remained within a 10 HU range in most patients. No significant attenuation differences were found between CI(art)/VMI(40keV) in arterial vessels (p-range 0.06–0.86). Contrast-to-noise ratio provided by VMI(40keV) and CI(art) was equivalent for all arterial vessels assessed (p-range 0.14–0.91). Qualitatively, VMI(40keV) showed similar enhancement of abdominal and pelvic arteries as CI(art) and VNC were rated comparable to TNC. CONCLUSION: Our study suggests that VNC and VMI(40keV) derived from single venous-phase SDCT offer comparable assessment of major abdominal vessels as provided by routine triphasic examinations, if no dynamic contrast information is required.