Purpose: To detect diminished renal function by extracting quantitative measures from changes in contrast media concentration in the kidney parenchyma and renal pelvis during multi-phase CT urography (CTU).
Methods: The degree of dilatation of the renal pelvis does not always correlate with the severity of renal obstruction. For assessing the severity of renal obstruction, an algorithm was developed to segment separately the renal parenchyma and renal pelvis of each kidney in each CTU phase and to evaluate the attenuation of the contrast media in both. Then, two quantitative indicators were calculated: the change in the total attenuation of contrast media in the renal parenchyma and pelvis, during the CTU phases and "T1/2", the drainage time from the renal parenchyma only. To study the performance of the algorithm, 15 patients who underwent CTU imaging were retrospectively collected. Visual inspection indicated a dilated renal pelvis with varying degrees of hydronephrosis in 6 kidneys, while 24 kidneys appeared normal.
Results: The algorithm successfully segmented the parenchyma and the renal pelvis with an average dice coefficient of 0.97. For both indicators, a statistically significant difference (p < 0.001) was found between the kidneys with dilated renal pelves and the normal kidneys. In the hydronephrotic kidneys the total attenuation of the contrast media did not decrease during the CTU examination. In all the hydronephrotic kidneys a low rate of kidney drainage was found with a T1/2 value longer than 20 minutes. In one case, both indicators implied low rate of kidney drainage while the renal pelvis appeared normal.
Conclusions: The automatic segmentation of the renal parenchyma and renal pelvis in all CTU phases provides additional, quantitative parameters for the estimation of the severity of renal obstruction which is not always correlated with the degree of dilatation.