OBJECTIVES: To measure and compare the objective image quality of true noncontrast (TNC) images with virtual noncontrast (VNC) images acquired by tin-filter-enhanced, dual-source, dual-energy computed tomography (DECT) of upper abdomen. MATERIALS AND METHODS: Sixty-three patients received unenhanced abdominal CT and enhanced abdominal DECT (100/140 kV with tin filter) in portal-venous phase. VNC images were calculated from the DECT datasets using commercially available software. The mean attenuation of relevant tissues and image quality were compared between the TNC and VNC images. Image quality was rated objectively by measuring image noise and the sharpness of object edges using custom-designed software. Measurements were compared using Student two-tailed t-test. Correlation coefficients for tissue attenuation measurements between TNC and VNC were calculated and the relative deviations were illustrated using Bland-Altman plots. RESULTS: Mean attenuation differences between TNC and VNC (HUTNC - HUVNC) image sets were as follows: right liver lobe -4.94 Hounsfield units (HU), left liver lobe -3.29 HU, vena cava -2.19 HU, spleen -7.46 HU, pancreas 1.29 HU, fat -11.14 HU, aorta 1.29 HU, bone marrow 36.83 HU (all P < .05); right kidney 0.46 HU, left kidney 0.56 HU, vena portae -0.48 HU and muscle -0.62 HU (nonsignificant). Good correlations between VNC and TNC series were observed for liver, vena portae, kidneys, pancreas, muscle and bone marrow (Pearson's correlation coefficient ≥0.75). Mean image noise was significantly higher in TNC images (P < .0001). Measurements of edge sharpness revealed no significant differences between VNC and TNC images (P = .19). CONCLUSION: The Hounsfield units in VNC images closely resemble TNC images in the majority of the organs of the upper abdomen (kidneys, liver, pancreas). In spleen and fat, Hounsfield numbers in VNC images are tend to be higher than in TNC images. VNC images show a low image noise and satisfactory edge sharpness. Other criteria of image quality and the depiction of certain lesions need to be evaluated additionally.
OBJECTIVES: To measure and compare the objective image quality of true noncontrast (TNC) images with virtual noncontrast (VNC) images acquired by tin-filter-enhanced, dual-source, dual-energy computed tomography (DECT) of upper abdomen. MATERIALS AND METHODS: Sixty-three patients received unenhanced abdominal CT and enhanced abdominal DECT (100/140 kV with tin filter) in portal-venous phase. VNC images were calculated from the DECT datasets using commercially available software. The mean attenuation of relevant tissues and image quality were compared between the TNC and VNC images. Image quality was rated objectively by measuring image noise and the sharpness of object edges using custom-designed software. Measurements were compared using Student two-tailed t-test. Correlation coefficients for tissue attenuation measurements between TNC and VNC were calculated and the relative deviations were illustrated using Bland-Altman plots. RESULTS: Mean attenuation differences between TNC and VNC (HUTNC - HUVNC) image sets were as follows: right liver lobe -4.94 Hounsfield units (HU), left liver lobe -3.29 HU, vena cava -2.19 HU, spleen -7.46 HU, pancreas 1.29 HU, fat -11.14 HU, aorta 1.29 HU, bone marrow 36.83 HU (all P < .05); right kidney 0.46 HU, left kidney 0.56 HU, vena portae -0.48 HU and muscle -0.62 HU (nonsignificant). Good correlations between VNC and TNC series were observed for liver, vena portae, kidneys, pancreas, muscle and bone marrow (Pearson's correlation coefficient ≥0.75). Mean image noise was significantly higher in TNC images (P < .0001). Measurements of edge sharpness revealed no significant differences between VNC and TNC images (P = .19). CONCLUSION: The Hounsfield units in VNC images closely resemble TNC images in the majority of the organs of the upper abdomen (kidneys, liver, pancreas). In spleen and fat, Hounsfield numbers in VNC images are tend to be higher than in TNC images. VNC images show a low image noise and satisfactory edge sharpness. Other criteria of image quality and the depiction of certain lesions need to be evaluated additionally.
Authors: Su Young Yun; Young Jin Heo; Hae Woong Jeong; Jin Wook Baek; Hye Jung Choo; Gi Won Shin; Sung Tae Kim; Young Gyun Jeong; Ji Young Lee; Hyun Seok Jung Journal: Neuroradiology Date: 2019-01-25 Impact factor: 2.804
Authors: Elizabeth George; Jeremy R Wortman; Urvi P Fulwadhva; Jennifer W Uyeda; Aaron D Sodickson Journal: Br J Radiol Date: 2017-09-22 Impact factor: 3.039
Authors: Julian L Wichmann; Pawel Majenka; Martin Beeres; Wolfgang Kromen; Boris Schulz; Stefan Wesarg; Ralf W Bauer; J Matthias Kerl; Tatjana Gruber-Rouh; Renate Hammerstingl; Thomas J Vogl; Thomas Lehnert Journal: Eur Radiol Date: 2014-07-17 Impact factor: 5.315
Authors: Ahmed M Tawfik; A A Razek; J Matthias Kerl; N E Nour-Eldin; Ralf Bauer; Thomas J Vogl Journal: Eur Radiol Date: 2013-10-02 Impact factor: 5.315