A technique for analysis of geometric mean renography

Abstract

Renography is used routinely to assess relative right to left renal function. Quantification is usually carried out using posterior images. Errors in relative renal function may occur if the kidneys are at different depths. Geometric mean images from combined anterior and posterior views are much less affected by kidney depth and offer the opportunity of more accurate and precise quantification. Background subtraction is a key part of the analysis process and validated protocols for geometric mean imaging have not been devised. This study aims to derive a suitable background subtraction protocol for geometric mean imaging. Simultaneous anterior and posterior renography using Tc mercaptoacetyltriglycine (MAG3) was performed on 16 adults. Analysis was carried out using both geometric mean and posterior images. The geometric mean background subtraction protocol was modified to give the same results as a posterior method, which had previously been validated by correlation with measurements of glomerular filtration rate. Absolute and relative uptakes were then obtained from both geometric mean and posterior analyses. For each analysis values were obtained both with and without depth correction. A revised background subtraction protocol for geometric mean renography was devised which operated successfully on all studies. Both absolute renal uptake and relative function values obtained from geometric mean analysis were not systematically different from those obtained using posterior analysis with depth correction. Values of the relative renal function from posterior analysis after depth correction were closer to the geometric mean values than estimates obtained before correction. A technique for analysing geometric mean renography data has been developed which gives results consistent with a previously validated posterior-only method.