A method for assessing split renal function in computed tomography: a retrospective study

INTRODUCTION: Accurate determination of the percentage of Renal contribution to their overall function is an important indicator when choosing clinical tactics for Renal transplantation and surgical interventions. The possibility of using nephroscintigraphy or complex alternative radiation methods is not always available. OBJECTIVE: Development and validation of methods for assessing split Renal function in a standard CT scan with contrast based on specialized automatic segmentation programs or manual measurement data. MATERIAL AND METHODS: The study included the results of a CT scan of the retroperitoneal space with contrast in 243 patients. 100 of them had nephroscintigraphy data. The Vitrea Advanced Visualization software was used for automatic segmentation of the renal parenchyma into the nephrographic phase. Data on volume ( V ) and average density ( HU ) were obtained. The values of renal parenchyma density in the native phase were performed manually. The split contribution of the Renals was determined based on the percentage ratio of the «mass» of the contrast agent in the parenchyma of each Renal, calculated as the product of the volume of the parenchyma and the average values of its density. The obtained results were compared with the nephroscintigraphy data. To develop a manual method for calculating the volume of the renal parenchyma, three linear Renal sizes (width, thickness, and length) and three parenchymal thicknesses were measured. The average values of renal parenchyma density were obtained based on three measurements of the round ROI at different levels. Statistics: The results obtained during the study were processed using statistical evaluation methods (correlation and covariance analysis). RESULTS: Using data on the volume values of the Renal parenchyma obtained during automatic segmentation using regression analysis, formulas for calculating the volume of the parenchyma were obtained based on taking into account all six linear Renal sizes with additional coefficients: V =1/1000(0.92 xyT –0.07 y 3 +0.07 yz 2 +0.19 z 2 T ), which demonstrated a high positive correlation and accuracy of the method (correlation coefficient — 0,99 (p