Fig. 2

A Scatterplot and linear regression of the relationship between the two methods of perfusion measurement. Values on the x-axis represent organ perfusion computed using the maximal upslope of the time-enhancement curve using all data points smoothed with a spline function. Values on the y-axis indicate organ perfusion interpolated using the maximal upslope of the gamma variate function using peak tissue enhancement and time-enhancement curve shape parameter (α) computed from the aortic curve. Regression equation: \(\mathrm{gamma}\;\mathrm{variate}\;\mathrm{based}\;\mathrm{perfusion}\;\left(\frac{\displaystyle\frac{\mathrm{ml}}\min}{\mathrm g}\right)=0.06+0.91\times\left(\mathrm{spline}\;\mathrm{based}\;\mathrm{perfusion}\right)\), r² = 0.92, p < 0.001. Three points for each patient are included: empty squares for the pancreas, empty triangles for the kidney, and black upside-down triangles for hepatocellular carcinoma (HCC). B Bland–Altman plot for the comparison of the two perfusion computation methods. The red line indicates the bias (-0.12), while green lines indicate the lower (-0.73) and upper (0.49) limit of agreement. Only one outlier was observed