Table 1.
Patient characteristics of each groups.
Fig 1.
Box-and-whisker plots of image noise of simulated 120-kVp SCM images and 40–70-keV nMERA HRCM images during HAP and PVP.
Boxes show the upper and lower quartiles, with median values being shown by the horizontal lines within the boxes. The whiskers represent maximum and minimum values. n.s. p > 0.05, * p < 0.05, ** p < 0.01.
Fig 2.
Box-and-whisker plots of the attenuation value and CNR of simulated 120-kVp SCM images and 40–70-keV nMERA HRCM images during HAP and PVP.
Boxes show the upper and lower quartiles, with median values being shown by the horizontal lines within the boxes. The whiskers represent maximum and minimum values. n.s. p > 0.05, * p < 0.05, ** p < 0.01.
Fig 3.
Box-and-whisker plots of the tumor to liver contrast and CNR of tumor in simulated 120-kVp SCM image and 40–70-keV nMERA HRCM images during HAP and PVP.
Boxes show the upper and lower quartiles, with median values being shown by the horizontal lines within the boxes. The whiskers represent maximum and minimum values. n.s. p > 0.05, * p < 0.05, ** p < 0.01.
Fig 4.
A representative case of a 68-year-old man with local tumor progression.
A 68-year-old man (body weight, 70.4 kg; BMI, 24.1 kg/m2) underwent multiphasic hepatic CT with dual source DECT at half reduced contrast medium dose. The image was reconstructed with nMERA images at 40–70-keV. All the images were under optimized window width/center setting to each image. The conspicuity of the hypervascular HCC (arrows) gradually improves as the energy level decreases in nMERA images, especially in the HAP image.
Table 2.
Qualitative image analysis.
Fig 5.
ROC curves of HRCM groups.
Fig 6.
A 64-year-old woman underwent CT for HCC one month before, one month and two months after radiofrequency ablation.
The image of HAP was reconstructed with nMERA HRCM images at 40–70-keV, and simulated 120-kVp SCM image. A high attenuating lesion abutting IVC was noted in nMERA 40–50 keV images. This lesion was not noted in 120-kVp images taken 1 month before and after the nMERA images, which indicating the false positive. All the images were under constant window width/center (300/40) setting.
Table 3.
Sensitivity, specificity, and AUC for detection of HCCs.