Background subtraction in 99Tcm-DTPA renography using multiple background regions: a comparison of methods

In order to obtain quantitative information from a 99Tcm-diethylenetriaminepentaacetate (DTPA) renogram it is first necessary to correct for the presence of background activity in the kidney time-activity curve. Three different methods of background subtraction which have been previously described are compared. Method 1 is a linear regression technique which uses separate tissue and heart time-activity curves to represent the background. Method 2 also uses multiple background regions and involves multiple regression. Method 3 uses a single heart region to represent the background. The three techniques are assessed using both simulation studies and patient data. The results show that method 3, which does not correct for extravascular activity, overestimates renal function particularly in poorly functioning kidneys. Method 1 gives the most accurate and reproducible results for both the simulation studies and patient data.