Table 1.
Clinical characteristics of the 96 patients.
Figure 1.
Selecting the best contrast-noise-ratio (CNR) for displaying gastric cancer with GSI Viewer analysis tool.
Optimal monochromatic energy of 70 keV achieved the best CNR for the primary lesion.
Figure 2.
Selecting the best contrast-noise-ratio (CNR) for displaying gastric cancer with GSI Viewer analysis tool.
ROI selections for primary lesion and normal gastric wall on an axial image.
Table 2.
Criteria for TNM staging by CT.
Table 3.
Accuracies, sensitivities and specificities for T staging using kVp images (Group A) and optimal monochromatic images (Group B) with histological examination as the reference standard.
Figure 3.
Monochromatic image in arterial phase of a 48-year-old man with stage T3N2 gastric adenocarcinoma demonstrated focal thickening (arrows) of the cardiac portion and lesser curvature with the abruptly interruption of mucous enhancement.
Figure 4.
Same patient as Figure 3.
KVp image in arterial phase maintained a relatively sharp tumor contour.
Figure 5.
Same patient as Figure 3.
Monochromatic image in portal phase demonstrated striation enhancement of blurring and wide reticular strands surrounding the outer border (arrow heads) of the tumor staged as T3 which was proved by histology.
Figure 6.
Same patient as Figure 3.
KVp image in portal phase maintained a relatively sharp tumor contour and a clear stomach fat plane (arrow heads), staged as T2.
Figure 7.
Same patient as Figure 3.
Five lymph nodes (arrow heads) were found in lesser curvature with monochromatic images.
Figure 8.
Same patient as Figure 3.
Four lymph nodes (arrow heads) were detected with kVp image.
Table 4.
Accuracies, sensitivities and specificities for N staging using kVp images (Group A) and optimal monochromatic images (Group B) with histological examination as the reference standard.
Table 5.
Accuracies, sensitivities and specificities for the distinction of N0 vs. N+ using kVp images (Group A) and optimal monochromatic images (Group B) with histological examination as the reference standard among patients with non metastatic disease.
Table 6.
Accuracies, sensitivities and specificities for M staging using kVp images (Group A) and optimal monochromatic images (Group B) with histological examination as the reference standard.
Figure 9.
Receiver operating characteristic curves for differentiating metastatic and non-metastatic lymph node in arterial phase.
Figure 10.
Receiver operating characteristic curves for differentiating metastatic and non-metastatic lymph node in portal phase.
Figure 11.
Water-based material decomposition images for a 62-year-old man with signet ring cell carcinoma obtained with dual energy spectral scan mode.
Figure 12.
Same patient as Figure 11.
Monochromatic image obtained at 70 keV energy level revealed the primary lesion (black arrows) and non-metastatic lymph node (arrow head) and metastatic lymph node (white arrow).
Figure 13.
Iodine-based material decomposition images of same patient as Figure 11.
The iodine concentration of the primary lesion (black arrows) was 37.77(mg/mL), non-metastatic lymph node (arrow head) was 26.00 mg/mL, and metastatic lymph node (white arrow) was 12.93 mg/ml.
Figure 14.
GSI scatterplot images water and iodine concentration plots for the same patient as Figure 11.
Table 7.
Comparison of nIC in AP and PP with clinic-pathological characteristics of the patients including histological classification and metastic or nonmetastatic lymph nodes.