Introduction. Assessment of separate renal function is an important diagnostic characteristic when planning kidney transplant operations or significant surgical interventions. Modern methods make it possible to obtain such data with the help of additional studies, in particular with nephroscintigraphy, which carries with it an additional financial burden and an increase in preoperative preparation time. The developed methods for assessing the separate function of the kidneys due to their automatic segmentation during CT require specialized workstations.
Purpose: development of a method for assessing separate renal function in a standard CT scan with contrast based on a new method for determining the volume of the renal parenchyma by individual linear dimensions and its densitometric parameters obtained by manual measurement.
Materials and methods. The results of 120 CT scans of the abdominal cavity and retroperitoneal space with contrast have been retrospectively analyzed. In order to obtain data on the "true" volume of the kidneys and their average density values in Hounsfield units, the kidneys were segmented into the nephrographic phase of contrast on a workstation using the Vitrea Advanced Visualization program. To further develop a rational approach to determining the methodology for estimating kidney volume based on manual measurements, three linear kidney sizes and three parenchymal thicknesses most used in practical work were selected and measured. Based on the data obtained, a regression analysis was performed with various combinations of its linear dimensions. To evaluate the optimal method for determining the densitometric parameters of the renal parenchyma, they were measured with a circular ROI on axial sections in three kidney sections. The average values obtained were compared with those obtained by tracing the kidney parenchyma in the frontal section. Based on the calculated data of the kidney volume and densitometric parameters of the renal parenchyma, the separate contribution of the kidneys to their overall function was calculated. The calculation results were compared with similar data obtained during kidney segmentation in the nephrographic phase of contrast on the workstation during their automatic segmentation.
Results. Regression analysis was used to obtain formulas for calculating parenchymal volume based on various combinations of its linear dimensions. Naturally, the formula that takes into account all six kidney sizes proved to be the most accurate: V=1/1000(0,92xyT-0,07y3+0,07yz2+0,19z2T). As a result of comparing two methods for estimating the average values of the densitometric parameters of the kidney parenchyma – using round ROI measurement and using manual circling – the round ROI measurement method was chosen, which showed the greatest correlation and covariance with similar data obtained with automatic kidney segmentation (the correlation coefficient was 0.988 (p<0.001); the covariance coefficient was 0.76 (p<0.001). The final analysis of the separate contribution of each kidney to their overall function based on the developed formula for manually calculating parenchymal volume and taking into account densitometric characteristics based on three measurements of parenchyma by round ROI demonstrated a high positive correlation with the results of calculating the separate contribution of each kidney obtained by automatic segmentation on a workstation. (correlation coefficient – 0.91 (p<0.001); covariance coefficient – 0.86 (p<0.001)).
Conclusions. The developed methods for manually determining the volume of the renal parenchyma and estimating the density characteristics of the renal parenchyma with round ROI showed a very high correlation with similar data obtained with automatic kidney segmentation. These methods are convenient in practical work and do not require special software tools. The combined use of the developed methods will in many cases avoid the need for additional nephroscintigraphy and reduce the time needed to prepare the patient for treatment.
