Renal volume assessment with 3D ultrasound

Abstract

This study was undertaken to determine the accuracy of 3D ultrasound (US) in assessing renal volume, with multislice computed tomography (MSCT) considered as the gold standard. Forty-nine patients (30 men, 19 women; age range 30-82 years) underwent abdominal contrast-enhanced MSCT and 3D-US performed with a 3.5-MHz 3D/4D convex-array probe. The results of the two modalities were compared with the Wilcoxon test. Variability between the two measurements was determined with the Bland-Altman method and reported in terms of bias and coefficient of repeatability (CoR). Mean values obtained were 210 ml with MSCT and 192 ml with 3D-US (p<0.001). Analysis of variability per patient between MSCT and 3D-US showed a bias of 19 ml, a CoR of 47 ml and an accuracy of 78%, with an average 3D-US underestimation of 19 ml (9%). Analysis of variability per kidney showed a bias of 9 ml, a CoR of 34 ml and an accuracy of 80%. Three-dimensional US is a valuable technique for monitoring renal volume, whereas MSCT may be reserved for assessing renal anatomy and relationships with neighbouring organs.