Image analysis
Liver and splenic attenuation measurements in HU were obtained by averaging two 1.0±0.1-cm2 circular region-of-interests (ROIs) placed above and below the main portal vein plane in both the unenhanced and portal venous phase contrast-enhanced images of each patient. To avoid partial-volume averaging effect, special care was taken to avoid measurements from vessels, focal lesions, areas of artifact, or near the edge of organ. Furthermore, only measurements from the posterior half of right hepatic lobe were obtained because of similar anatomic location between liver and spleen with resultant similar artifacts (Figure 1).
Liver and splenic attenuation measurements were averaged two measured levels to calculate the mean attenuation values for each patient. Liver minus splenic (L-S) attenuation values were …show more content…
Biopsy proof was not obtained.
Statistical analysis
The optimal threshold for diagnosis of hepatic steatosis on contrast-enhanced CT was determined by Receiver Operator Characteristic (ROC) curve and calculated sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy over a range of possible liver attenuation diagnostic threshold, L-S attenuation diagnostic threshold or combined. The maximal overall accuracy value was chosen.
Presence of focal fat sparing was evaluated efficacy for diagnosis of hepatic steatosis by calculating sensitivity, specificity, PPV, NPV and accuracy.
To evaluate the effect of delayed CT scan time, rate of contrast injection, volume and concentration of contrast media on hepatic and splenic enhancement as well as L-S attenuation that should affect the accuracy for diagnosing hepatic steatosis, Pearson correlation was used for
Liver and splenic attenuation measurements in HU were obtained by averaging two 1.0±0.1-cm2 circular region-of-interests (ROIs) placed above and below the main portal vein plane in both the unenhanced and portal venous phase contrast-enhanced images of each patient. To avoid partial-volume averaging effect, special care was taken to avoid measurements from vessels, focal lesions, areas of artifact, or near the edge of organ. Furthermore, only measurements from the posterior half of right hepatic lobe were obtained because of similar anatomic location between liver and spleen with resultant similar artifacts (Figure 1).
Liver and splenic attenuation measurements were averaged two measured levels to calculate the mean attenuation values for each patient. Liver minus splenic (L-S) attenuation values were …show more content…
Biopsy proof was not obtained.
Statistical analysis
The optimal threshold for diagnosis of hepatic steatosis on contrast-enhanced CT was determined by Receiver Operator Characteristic (ROC) curve and calculated sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy over a range of possible liver attenuation diagnostic threshold, L-S attenuation diagnostic threshold or combined. The maximal overall accuracy value was chosen.
Presence of focal fat sparing was evaluated efficacy for diagnosis of hepatic steatosis by calculating sensitivity, specificity, PPV, NPV and accuracy.
To evaluate the effect of delayed CT scan time, rate of contrast injection, volume and concentration of contrast media on hepatic and splenic enhancement as well as L-S attenuation that should affect the accuracy for diagnosing hepatic steatosis, Pearson correlation was used for